200 Glades Rd., Boca Raton, FL 33432
World Journal of Stem Cells

Bone marrow-derived products: A classification proposal – bone marrow aspirate, bone marrow aspirate concentrate or hybrid?

Joseph Purita, José Fábio Santos Duarte Lana, Morey Kolber, Bruno Lima Rodrigues, Tomas Mosaner, Gabriel Silva Santos, Carolina Caliari-Oliveira, Stephany Cares Huber

Download the PDF


Degenerative musculoskeletal disorders are one of the top causes of pain and disability in the adult population. Current available alternatives to mitigate symptoms include conservative treatments such as the administration of pharmacological agents and an educative approach towards lifestyle modification. The use of certain analgesics, such as opiates and corticosteroids, delivers short term results but do not address the etiological source of pain and disability. Also, prolonged use of such medications may cause additional complications. Therefore, the demand for musculoskeletal tissue regeneration has led to an alternative approach referred to as “orthobiologics”. This alternative is based on cellular and molecular components capable of inducing and promoting tissue repair. Bone marrow (BM) aspirate (BMA) and concentrate are well-known orthobiologics used to treat musculoskeletal conditions. Orthobiologics derived from the BM have been discussed in the literature; however, the lack of standardization regarding collection and processing protocols presents a challenge for generalization of study outcomes and determination of efficacy. Since BM-derived orthobiologics have not yet been classified, to our knowledge, this manuscript proposes the ACH classification system, which speaks to BMA (A), BMA and concentrate (C) and hybrid (H), which combines A and C. This classification proposes and describes 8 parameters that are relevant for the quality of biological products. The more parameters used would imply greater characterization and complexity of the evaluation of the biological product used. The ACH classification envisages a necessary contribution to the comprehension of both clinical procedures and research outcomes, ultimately ushering in a standardization of best practice.

Core Tip: Degenerative musculoskeletal disorders are one of the top causes of pain and disability in the adult population. The use of certain analgesics delivers short term results but do not address the etiological source of pain and disability. The demand for musculoskeletal tissue regeneration has led to an alternative approach referred to as orthobiologics, which is based on cellular and molecular components capable of promoting tissue repair. Bone marrow aspirate and concentrate are well-known orthobiologics used to treat musculoskeletal conditions. Since bone marrow-derived orthobiologics have not yet been classified, to our knowledge, this manuscript proposes the ACH classification system.


The increasing incidence of degenerative diseases affecting the musculoskeletal system is the main cause of pain and disability among adults. Current options for the management of these conditions mainly focus on conservative care such as activity modification and pharmacological therapies. While pharmacological therapies such as opiates and non-steroidal anti-inflammatory drugs (NSAIDs) or corticosteroids offer short term efficacy, they are associated with well-known side effects if used on a longterm basis[1,2] . Moreover, few options exist outside of surgical solutions for those individuals recalcitrant to conservative care. The need for musculoskeletal tissue regeneration has led to an alternative approach referred to as orthobiologics, which is based on cellular and molecular components responsible for inducing and promoting tissue repair[3] . Orthobiologics, which comprise platelet rich plasma (PRP), bone marrow (BM) aspirate (BMA) and concentrate (BMAC), fat grafting (Bio fat), and expanded mesenchymal stem cells (MSCs), have shown promising results for the care of musculoskeletal disorders[4-7] . Orthobiologics have been discussed in the literature with promising results, however, the lack of standardization regarding the methods of obtaining and processing the cells and associated components, have led to uncertain conclusions in terms of efficacy and ability to generalize outcomes[8] . Specifically, the main components of orthobiologics (platelet concentrations, growth factors, and cytokines) may vary based on the processing method, which might affect anabolic and antiinflammatory properties, and consequently lead to inconsistent outcomes[8] . Thus, the need for standardization and classification of orthobiologics is imperative for understanding procedures and dissemination of research outcomes. A classification system has been developed for PRP[9] ; however, no such classification exists for BMderived orthobiologics. Thus, the purpose of this paper is to present a proposal for a classification system for BM derived orthobiologics.


The main function of BM is to provide circulating blood with an optimal supply of erythrocytes, leukocytes, and platelets. In addition to this, BM supplies hematopoietic stem cells (HSCs), endothelial cells, MSCs and other precursor cells. The human skeleton possesses red BM which is hematopoietically active, and yellow, which is hematopoietically inactive[10] . Red and yellow BMs have different cellular and molecular content: Yellow BM comprises 95% fat cells, whereas the red BM comprises 60% hematopoietic cells. The whole skeleton is filled with red BM at birth, however, during childhood a physiological conversion of red BM into yellow BM occurs. The conversion of red to yellow marrow and progresses to the axial skeleton, and this entire process may be completed by the age of 25 years[10] . BM is a potent source of stem and progenitor cells, and this characteristic has gained attention for cell-based therapies in orthopedics[11-13] . Given the diversity in stem cell lineages and phenotypes in the marrow, BM represents a functional organ in which distinct types of cells function cooperatively. Specifically, HSCs play a critical role in the formation of the hematopoietic microenvironment, whereas MSCs support hematopoiesis and both MSCs and/or skeletal stem cells are responsible for the development and maintenance of skeletal tissues.

Cellular Content

MSCs are non-hematopoietic stromal cells that are composed of a small fraction (0.001%–0.01%) of the stem cell content in BM[16] . MSCs are found in other tissues, such as adipose tissue, placenta, and umbilical cord, and although they differ in their differentiation potential, they possess common features associated with those from the BM, which might imply that MSC-like populations share a similar ontogeny[17,18] . MSCs exhibit the potential ability to differentiate into mesodermal linage cells (e.g., cartilage, bone, fat, muscle, meniscus and tendon)[19] , which is fundamental for the regeneration process. Moreover, these cells have paracrine effects, thus are able to alter their local microenvironment[20] . Given the varying MSC markers that laboratories may use to characterize these cells, there is a lack in standard phenotypic criteria. This heterogeneity is also due to the fact that MSCs are able to express a range of cell-lineage specific antigens that may differ depending on the culture preparation, culture duration, or plating density[21,22] . However, the Mesenchymal and Tissue Stem Cell Committee of the International Society for Cellular Therapy have proposed minimal criteria to characterize MSCs, which comprise the following attributes: Must be plastic-adherent when maintained in culture; must be able to differentiate in vitro into chondroblasts, adipocytes and osteoblasts; and must express CD105, CD73 and CD90, and lack expression of CD45, CD11b, CD34 or CD14, CD79α or CD19 and HLA-DR surface molecules[23] . MSCs lack significant immunogenicity and can be easily isolated, which allows allogenic transplantation. In allogenic circumstances these cells should be considered immune evasive. However, the effects of MSCs in cellular-based therapies depends on the ability of these cells to home and engraft (long-term) into the target tissue[24] . One theory suggests that MSCs have a rather short life span and are phagocytized by monocytes and subsequently stimulate the production of T-reg cells which may very well contribute to the overall clinical improvement[25] . Cells from injured tissue release chemokines responsible for MSC recruitment. Once in the target tissue, MSCs are able to modulate wound-healing responses by reducing apoptosis and fibrosis, attenuate the inflammatory process and stimulate cell proliferation and differentiation via paracrine and autocrine pathways[26] . These properties are attributed to the ability of MSCs to release key agents, such as vasculoendothelial growth factor, transforming growth factor beta (TGF-β), stromalderived factor 1, and stem cell factor, among others. Also, they induce a downregulation of pro-inflammatory cytokines, including interleukin 1 (IL-1), IL-6, interferon-γ, and tumor necrosis factor α[16,27,28] . MSCs also possess immunomodulatory properties as they are able to inhibit the activation of type 1 macrophages, natural killer cells, and both B and T lymphocytes[29] . HSCs also known for expressing CD34+, are located at the top of the hematopoietic hierarchy. They are responsible for the daily supply of more than 100 billion mature blood cells, including erythrocytes, leukocytes, and platelets[30] . This process, called hematopoiesis, is of extreme importance in the maintenance and regulation of the immune system, especially for the cells from myeloid lineage, such as granulocytes, monocytes and dendritic cells, due to their short half-life[31,32] . Past studies have reported that the hematopoietic and stromal environments are related and overlapped. For example, Simons et al[33] observed generations of fibroblasts colony-forming unit (CFU-F) from CD34+ human BM cells. Also, it has been reported that the number of osteoblast progenitor cells is higher in sorted CD34+ cells (1/5000 approximately) than in CD34 - populations (1/33000), and when these sorted cells were cultured in a long-term marrow system, the generation of a heterogeneous population that included smooth muscle cells, adipocytes, fibroblast and macrophages was observed[34] . This possible relation was then supported by Mehrotra et al[35] who reported that HSC give rise to osteocytes and chondrocytes in an experimental study. Immune cells – Leukocytes have a common origin from the hematopoietic stem cell and develop along distinct differentiation pathways in response to external and internal stimuli. In order to promote regeneration, leukocytes circulate through the blood and lymphatic system and are recruited to specific regions of the body when damage occurs[32] . The mononuclear phagocyte system represents a subset of leukocytes that was originally described as BM-derived myeloid cells[32] . Monocytes are immune effector cells that, although they circulate in the blood, BM and spleen, they do not proliferate in a steady state[36,37] . They are equipped with chemokine receptors that mediate migration from blood to the injured sites and produce inflammatory cytokines. During inflammation, the monocytes differentiate into dendritic cells (DC) or macrophages, and this process is likely determined by the inflammatory environment and pathogen-associated pattern-recognition receptors[38] . Macrophages are phagocytic cells that reside in lymphoid and nonlymphoid tissues[32] . Given the broad range of pathogen-recognition receptors that macrophages possess, they are known as an efficient tool at maintaining tissue homeostasis as they provide clearance of apoptotic cells and remodeling of the extracellular matrix[39,40] . Macrophages play a key role in recruiting and inducing the proliferation of osteoblasts, stem and progenitor cells as they secrete bone morphogenetic proteins, IL-1β, TGF-β, platelet derived growth factor and insulin-like growth factors, in areas of infection or injury in different tissues in the body[41] . Extrinsic stimuli that induce an inflammatory process, such as infection or injury, promote changes in gene transcription that classify macrophages as type 1 (M-1) and type 2 (M-2). The M-2 type offers a healing function, while M-1 promotes the host defense. After injury, M2 can switch into M1, and this change is modulated by the cytokines such as interferonγ, and M2 type by IL-4[42] . Neutrophils belong to a polymorphonuclear family and are known for being the main cell type response to bacterial infections. It was reported that neutrophils are highly plastic cells influenced by environmental cues that result in a site-specific neutrophil transcriptome as they migrate from BM to sites of inflammation[43] . As a granulocyte, which includes eosinophils and basophils, neutrophils are able to secrete a variety of cytokines, such as TGF-β, vasculoendothelial growth factor and platelet derived growth factor, playing an important role in angiogenesis and vasculogenesis[44] . Neutrophils undergo spontaneous apoptosis to regulate the resolution of inflammation[45] .



The main goal in treating orthopedic injuries, especially joint disease, is cartilage regeneration. One approach to achieve this outcome is by using BM-derived MSC (BM-MSC), which has been supported in the literature[46,47] . However, its clinical utility is limited by complexity, such as the need for a specialized laboratory and procedural cost. In this sense, the use of BMA has emerged as a novel regenerative tool for degenerative joint diseases as a non-fractioned product that retains potentially supportive chondrogenic components[48] . Even though different harvest sources for BM have been described in the literature the main harvest site (either for BMA or BMAC use) is the posterior iliac crest, which allows a considerable amount of BM and about 1.6-fold more osteoblastic connective tissue progenitor cells than other sites[49,50] . However, evidence suggests the quality of the product is technique-dependent[51] . There are a few studies that have used this approach in the literature; however, most of them are related to nonunion fractures. The first to describe the use of unprocessed marrow was Lindholm and Urist[52] that reported the replacement of bone matrix by new bone in composite grafts in vivo (non-human study). Almost a decade later, Connolly et al[53] observed callus formation sufficient to unite tibial nonunions in humans after injection of autologous BMA. In 2013, Hauser and Orlofsky published a case series describing their experience with BMA in combination with hyperosmotic dextrose, also known as prolotherapy, in the treatment of knee, hip, and ankle osteoarthritis. After two to seven treatments over twelve months, all patients reported improvement in pain, joint function, and quality of life. Also, three out of seven patients had achieved complete symptomatic relief[48] . Butala et al[54] reported the efficacy of BMA in bone union as they injected unprocessed BM at fracture sites in 10 patients with tibia, humerus, femur, and forearm delayed union fractures. After 12 wk, nine of these patients had signs of union, such as decreased tenderness at fracture site, pain-free joint mobilization and ability to ambulate without assistance[54] . A study performed in 2017 by Lal[55] , evaluated the use of percutaneous autologous BM injections in 56 patients with delayed and 37 patients with nonunion of long bone. Twelve weeks after the injections, it was observed that all fractures were united, and the minimum period for union was 8-weeks. Although a significant correlation (P = 0.081) was not present, it was reported that the time to observe bone union after the injection of autologous BM was longer in patients who were smokers. Women, however, were observed to have a reduced time for bone union than the male patients (P = 0.041)[55] . Although the number of studies with BMA are limited and of lower quality, they show a promising efficacy and safety profile with regards to adverse events.


In an attempt to increase the proportion of MSCs, the aspirate of BM may be processed to produce BMAC, which has been widely investigated in orthopedics, especially for nonunions, surgical augmentation, osteonecrosis, as well as osseous and cartilage defects[11-13] . Although the exact mechanism of action has not been fully elucidated, the effects of BMAC may rely on the recovery of nucleated cells from BM, which possesses a paracrine effect by delivering cytokines into the injured site in order to stimulate endogenous tissue repair[56] . In vitro studies have shown that the platelets present in BMAC release growth factors that induce stem cells migration to the injured area. Moreover, a concentrated number of HSCs may provide vascular support and drive MSC into osteogenic differentiation pathways[57] . Current clinical studies have reported the efficacy and safety of BMAC for the treatment of small lesions. Centeno et al[6] studied the effects of BMAC on 115 shoulders of 102 patients who had rotated cuff injuries and shoulder osteoarthritis. In the aforementioned study, a 52.6% improvement in joint function and disability and 44.2% decrease in pain was reported with both outcomes reaching statistical significance (P = 0.001). The mean improvement reported by the patients was 48.8%. The reduction of disability and pain was observed from the first month after treatment and was maintained for up to 2 years after the treatment, based on this time being the terminal point of data collection. No side effects or adverse events were reported with BMAC in these 2 years of study[6] . BMAC has also been studied with various surgical scaffolds. Gobbi et al[58] evaluated 15 patients with grade IV cartilage lesions who underwent injections of BMAC on a collagen matrix. Two years after the injections improvements in pain, joint functionality and quality of life were identified. Biopsy of these lesions showed hyaline-like tissue at repeat arthroscopy 2- years later[58] . Enea et al[59] evaluated patients who underwent microfracture covered with a resorbable composite of natural hyaluronan matrix and synthetic polyglycolic acid with BMAC. It was observed that, 12 mo after the injection, the lesions were macroscopically normal, presenting production of hyaline-like tissue. The defect filling was confirmed by magnetic resonance imaging[59] . The use of BMAC has also been studied in combination with other regenerative medicine approaches. Sampson et al[60] evaluated the injection of BMAC followed by PRP in 125 patients who presented moderate/severe ankle, knee, spine and/or shoulder osteoarthritis, eight weeks after the injection, The authors observed a median of 5 points in pain relief, based on a visual analogic scale (VAS), and the patients reported 9.0/10 satisfaction with the treatment. Kim et al[61] studied the association of BMAC with adipose tissue (fat graft) in 75 osteoarthritic knees (41 patients). Twelve months after the injections, a decrease in pain, improved joint function, and an increase in quality of life was reported. The authors also suggest that BMAC would present a more effective result in early to moderate phases of osteoarthritis. Some studies evaluated the optimal volume of BM needed to achieve clinical response: The quality of the product decreases with higher volume of BM withdrawn, and it was observed that small volume of marrow aspirated in a 10 mL syringe would be an ideal volume to concentrate MSC and progenitor cells. Larger volume syringes may cause blood dilution[62,63] . The components of BM aspirated are concentrated following centrifugation steps. Although there are some protocols of BMAC preparation in the literature[64,65] there is no study regarding the optimal centrifuge force and time to achieve an increased cellular concentration. Although BMAC presents a well-established cellular and molecular content, only few studies evaluating its efficacy and safety have performed quantitative and qualitative assessment[8]


The lack of standardization of the BM-derived products for regenerative medicine has emerged, thus the need to classify the processing methods according to quality and procedural details has been established[11] . Classification of such factors would allow for procedural standardization and interpretation of both clinical results and research findings. The ACH (aspirate, concentrate, hybrid) classification system comprises the two main techniques involving bone marrow-derived products: BMA, which represents the letter A (for aspirate), BMAC, which represents the letter C (for concentrated), and the letter H (for hybrid) is used when BMA is combined with BMAC. The ACH classification is focused on whether the cellular and molecular content present in the product was evaluated and described increasing the complexity of description/characterization. For each classification (A, C and H) sub grouping would occur as follows: (1) Product would only be collected and injected with no additional analysis; (2) Description of harvesting – BM site of harvesting (posterior/anterior iliac crest, axial skeleton), type of needle, multiple insertions, single insertion, type of syringe, type of anticoagulant, volume harvested; (3) The cellular content would be assessed by a cell count machine, which would enable to quantify mono- and polymorphonuclear cells, giving the number of total nucleated cells; (4) Dosage of molecular content, such as interleukins and/or growth factors is made by multiplex platform or ELISA technique; (5) Indirect quantification of MSC number measured through CFU in culture; (6) Phenotyping of MSC and HSC for characterization through flow cytometry – it is wise to use a full panel for the clusters of differentiation, especially of the MSC since there are a lot of markers for positive and negative evaluation; (7) For the complete characterization of MSC the differentiation in three cell types in culture is necessary, including the induction of chondrocytes, adipocytes and osteocytes; and (8) To finalize, the most complex level of evaluation of MSC is the evaluation of its function, which includes assays like wound healing (proliferation and migration), lymphocytes proliferation (immunossupressor potential), and population doubling time. The representation of the ACH classification is shown in Table 1. The idea of this classification is that for each type of BM used (BMA, BMAC or hybrid) the increase of the number indicates an improvement in the characterization and complexity of the evaluation of this biological product. When a study or procedure with BMAC reports that only BM was collected and injected, it would be classified as C1, according to the ACH classification. On the other hand, if the BMAC presents the description of harvesting procedure (site, syringe, volume, and anticoagulant use) it will be classified as C1-2. If the total nuclear cells were counted by a cell counter, which would include leukocytes, MSC and HSC, using the description of technique for harvesting it would be classified as C1-3. In this BMAC if the harvesting technique was described, cell count was made and evaluation of molecular content, it will be classified as C1-4. If the HSC and/or MSC are quantified and characterized by flow cytometry in the same BMAC, has the description of harvesting, dosage of cytokines and CFU it would be classified as C1-6 product. The last level of description is the C1-8 which encompass the description of harvesting, cell count, evaluation of cytokines and growth factors, MSC and HSC phenotyping, CFU, evaluation of differentiation and functional assays, being classified as C1-8. In the case where this is not a progression of steps in numeric order the omitted step number would not be used. For example, if a procedure with BMA harvesting had a description, cell count, and CFU, without the quantification of the molecular content, this study will be classified as A1-3;5, as demonstrated in Table 2. For a general view of ACH, we described a schematic illustration of the ACH classification exemplified by Figure 1.


Although studies using both BMA and BMAC for the treatment of various musculoskeletal disorders have shown promising clinical results, inconsistent preparation methods with deficient reporting has led to questionable outcomes with respect to generalization and reproducibility. In order to optimize the efficacy and safety of BM-derived products, and to allow validation and standardization of such products, studies should report stepwise descriptions of the preparation protocol and additional information to further classify the product used. The ACH classification focuses on describing parameters that are relevant for the quality of biological products, such as the collection technique, cell count and its nature (whether stromal or hematopoietic), and molecular content dose. The ACH classification would contribute to a greater understanding of both clinical procedures and research outcomes and, over time, lead to a standardization of best practice. Together, we believe that the ACH Classification proposal is an easily recalled and useful method for the classification of BM-derived products in order to provide a comparative between product composition and clinical outcomes. It should also be emphasized that this classification is pertaining only to BMA products. There are other aspects of bone marrow preparation such as photobiomodulation of the aspirate or the concentrate that have not been discussed. Unfortunately, there is not much literature supporting this concept. Thus, this is mentioned as a matter of anecdotal interest.


Immediately following a procedure, it is common to experience swelling, erythema, and increased pain (16). Furthermore, many individuals who receive a PRP injection have had clinical signs and symptoms for a variable period before the injection. Moreover, most individuals have been recalcitrant to previous conservative treatments. Thus, the patient will not be pain-free after the injection and may have increased pain and swelling until the potential PRP-induced inflammatory symptoms resolve. At the early postprocedural stage, protection of the treated area, active rest, and management of the acute symptoms are the goal (41,61). Patients are encouraged to resume use of previously prescribed protective measures such as bracing or ambulatory aides. The decision to immobilize a joint is individualized and based on injury severity and location, as well as the patient’s typical activities of daily living. In the early acute stage, the physician administering the PRP injection will provide specific instructions for immediate and early care. Given the absence of specific guidelines from the literature, material presented in this section will be largely drawn from the authors’ experience, individual outcome–based studies, biological evidence, and previously published narrative discussion articles (3,21,28,41,51,75,80). Nevertheless, the physician who provided the injection should be consulted before engaging in any activities at this stage. Generally, in the early acute stage, the joint is protected for up to 3 days. Protection may be in the form of partial weight-bearing, use of sling or motionrestricted walking boot, or specific instructions to avoid use. Complete rest is discouraged, and active movements of the affected body part should be performed a few times a day to avoid the deleterious cellular effects of immobilization. The decision to limit weight-bearing is based on the patient’s presentation before the injection and the diagnosis. For example, a PRP injection to the plantar fascia may interfere with pain-free ambulation for a few days leading to the use of a restricted motion walking boot or partial weight-bearing. After the third day, a gradual increase in activity is encouraged, provided movements are tolerable (pain during movement that ceases after completion). Individuals with restricted weight-bearing are encouraged to progressively increase walking (duration and distance), and a progressive loading program is initiated. Failure to move the involved body part or weight-bear through the extremity may result in impairments that perpetuate or compound the pathology being treated. Similarly, overzealous activity will be counterproductive and may prolong the acute inflammatory phase. During the early and late acute stages, patients may be provided with medications to reduce pain and interfere with inflammation. Although there is no consensus on the use of medication, approximately 42% of physicians limit the use of NSAIDs in some manner, despite limited evidence of these medications interfering with PRP outcomes (61,80). Alternative medications that may be used include narcotics (e.g., Percocet) and acetaminophen (e.g., Tylenol) (80). Corticosteroids are strictly discouraged because they are associated with reduced chondrocyte and tenocyte proliferation as well as cellular apoptosis (11,17,68). Furthermore, corticosteroids are likely to interfere with the inflammatory phase, which may preclude regeneration. Moreover, in vitro evidence suggests that adding a corticosteroid to a PRP preparation may reduce cellular proliferation (17). Regardless of the evidence, patients or clients should consult their physician before taking any medications. Cryotherapy (e.g., ice) may be used to reduce pain and inflammation; however, a concern is that ice may reduce platelet activation and interfere with the intended inflammatory cascade after PRP (61,80). Regarding ice, evidence suggests that hypothermia offers a protective effect of cartilage after trauma (76) and may reduce subacromial bursal thickness (72). Moreover, there is no published evidence that implicates ice as a factor that may hinder PRP outcomes. Given that no study has demonstrated an inferior result from ice and studies have used ice after a PRP injection with positive outcomes (51,75), there is no conclusive evidence to preclude use.


After the early protective stage, stretching and isometrics are introduced at the affected regions. Isometrics have been shown to beneficially affect tendon healing, and evidence suggests that unloading of muscle activity (through Botox) impairs the regenerative effects of platelets (44,83). In cases where contractile tissue is being treated (e.g., tendon and muscle), the isometric efforts are submaximal. If the treated pathology is noncontractile tissue such as ligament or capsule, then the isometric contractions may be maximal effort. Isometrics should be repeated multiple times per day with 10-second holds working toward longer durations of up to 2 minutes by the subacute phase if tolerated. Activities such as cycling, deep-water running or water walking, and upperbody ergometry are introduced and advanced based on tolerance. In cases where a PRP injection is administered to treat OA, progressive and repetitive loading of the joint is encouraged. Activities such as low-resistance cycling and high-repetition heel slides (Figure 1) should be performed multiple times a day. A progressive increase in overall activities is pursued until the conclusion of week 2, when more pronounced advancements could be made. Before progression into the next phase, symptom tolerance to exercises must be gauged. The presence of some discomfort or pain during exercises is generally acceptable, provided symptom resolution occurs on completion. Pain or discomfort that occurs during exercises and remains elevated on completion suggests the need to evaluate and potentially regress programming.


The subacute phase is a key milestone, as this is where the preinjection heterogeneity has a considerable influence. Specifically, the variability of clinical presentation before the PRP procedure will influence progression. Although those who were functioning at a high level (sport participation or weight lifting) before injection will be placed on a more progressive routine, others will follow a more conservative approach. In consideration of safety, a more conservative approach is presented. The objective at this phase is to ensure that the cointroduction of mechanical forces is a primary component of the management strategy. The introduction of new exercises (and advancement of previously prescribed activities) that have been associated with regenerative changes in pathological tissue, upregulation of growth factors, and stimulation of cellular tissues is the overarching goal. Tables 1 and 2 highlight the associated exercises and progressions. At week 3, isometric exercises are advanced to full effort for all conditions and increased hold times are recommended (full or reduced effort to failure if tolerated). Although isometrics may be considered easy for some, continuation is recommended, given the biological benefits (anabolic) previously discussed. Eccentric-focused muscle actions are first introduced within a limited tissue range and ultimately advanced to full range consistent with published protocols (27,51,73). Studies using eccentric loading after PRP range in performance frequency from 3 times a week to twice daily (2,19,21,26,51,73,79). The authors of this article generally base frequency on premorbid conditioning level with a frequency of 3 times a week up to once daily. Isotonic strengthening is introduced based on impairments, and joint loading is gradually increased from protected to full weight-bearing. An elliptical trainer may be used for lower extremity pathology, and squatting on an incline device to unload the joints is encouraged (e.g., Total Gym), particularly in cases of OA or cartilage pathology. In some published investigations, individuals with focal tendinopathy or muscle strain have been progressed in a more aggressive manner after a PRP injection. For example, a case-control study of professional football players with a grade 2 (average grade) hamstring strain reported initiating a stretching program immediately and commencement of resistance training within the first week (75). In another study, patients with patellar tendinosis began a progressive eccentric loading program performed 3 times a week at the onset of week 2 (51). In the study, eccentric loading of the patellar tendon was gradually progressed with knee flexion limited to 458 for 1 week, which was progressed to 608 until week 6. At week 6, flexion was progressed to 908 with a decline board initiated shortly afterward (Figure 2). In another study of patellar tendinopathy, athletic participants were allowed to resume premorbid activities within a week after their first PRP injection (32). Boesen et al. (14) initiated a twice-daily eccentric loading program and allowed 5 minutes of running within the first week after PRP injection for patients with Achilles tendinopathy. In contrast to the aforementioned studies, a case report of a patient with a partial distal triceps tear used 2 weeks of rest before any exercise-based intervention and initiated eccentric loading of the triceps at week 6 (21). The above studies highlight the variability of programs and the need to individualize progressions based on symptoms and underlying injury.


The remodeling phase begins approximately at week 6 or 7. It is at this time when more advanced loading is initiated. Eccentric overload exercises are progressed to full range and heavy slow concentric actions are emphasized, with efforts to promote further soft-tissue remodeling. Although some examples of eccentric loading are provided in this article, readers may consult previously published articles in the Strength and Conditioning Journal for additional examples (18,58,50). Depending on the underlying injury, running is generally introduced, and individuals who are more active may be progressed to plyometric training with efforts to resume premorbid activities. Although running may be considered a risk factor for musculoskeletal conditions, much of this is speculative because recreational running is not a risk factor for lower extremity OA, and there is a body of evidence that it may positively increase the composition (e.g., hydration, proteoglycan content, and fiber hypertrophy) of the intervertebral disc (12,56). Individuals with OA are progressed to an appropriate exercise routine that promotes full range of motion with variable loading patterns based on clinical presentation. It is important to note that many of the activities introduced in this phase may be commenced earlier in athletic individuals who were functioning at a high level before injection. In keeping with a “one size does not fit all” approach, individuals who were functioning at a much lower level (e.g., potential surgical candidate, use of assistive device for walking, or advanced degenerative disease) before injection may not reach more advanced performance milestones such as treadmill or plyometric training.


The indications for a PRP injection are wide ranging and inclusive of OA, tendinopathy, cartilage defects, and muscle or tendon tears. Although each of these conditions (and the individual affected) have their own clinical presentation, degree of healing after the onset, and characteristic response to PRP, overarching considerations exist and should be used to narrate the postprocedural phases. Musculoskeletal injuries or disorders present a constellation of signs and symptoms, with no paucity of evidence for rehabilitation or postrehabilitation guidelines because resumption of premorbid function is the goal. After a PRP injection, these goals remain; however, there are additional considerations for which this column sought to address, namely, the protective phase where pain and inflammation are addressed, as well as the subacute and remodeling phases where exercise principles are used to augment the effects of PRP.


In the early acute phase, evidence from numerous studies and narrative articles suggest a short period of pain and inflammation. Although there is a consensus for a few days of rest, variability exists regarding the use of NSAIDs. The concern for NSAID use after PRP is based on reduced platelet aggregation, impaired growth factor release, interruption of the intended inflammation, and deleterious cellular effects. Regarding cellular function, the evidence is divergent. In one study of individuals exposed to a prolonged bout of running, NSAID use abolished the adaptive rise in collagen synthesis seen in the subjects who took a placebo (22). In another study, NSAIDs were infused into the quadriceps of subjects before, during, and for 4.5 hours after an eccentric training bout (63). Results indicated that the NSAIDs prevented satellite cell activity in the infused quadriceps, whereas the control extremity showed a 96% increase in activity. In contrast to the aforementioned study, Mackey et al. (60) studied the effect of NSAID ingestion on satellite cell activity of the quadriceps muscle (through biopsy) before and after (up to 30 days) a laboratoryinduced muscle injury. In the study, NSAID consumption was associated with increased satellite cell activation compared with placebo, suggesting that NSAIDs may have a different effect on injured versus healthy muscle (60). Additional evidence on the effect of NSAIDs may come from clinical studies. In one study, individuals with chronic Achilles tendinopathy were randomized to 1 week of NSAIDs or placebo (42). Results from biopsy and ultrasound indicate that ibuprofen had no effect on collagen expression and growth factors. Dideriksen et al. (30) studied the effects of NSAID use on patellar tendons of elderly adults who were immobilized for 2 weeks. Although collagen synthesis was decreased in the immobilized limbs as expected, the NSAID group revealed no difference when compared with placebo. Although evidence supports the premise that nonselective NSAIDs (e.g., naproxen) may interfere with platelet aggregation (55), growth factor levels in PRP are not altered (59,82). In conclusion, NSAID use does not compromise growth factor release and is more likely to have a deleterious effect on healthy structures as opposed to pathological tissues. Moreover, other medication classes such as acetaminophen and select opioid drugs may negatively influence postexercise protein synthesis and produce in vitro chondrotoxicity, respectively (1,81). Given the potential for NSAIDs to reduce pain and inflammation and the potential of other medication classes to negatively influence cells, it seems there is little downside to intermittent, low-dose use in the early acute phase


Formal postprocedural management strategies have been documented in the literature after PRP injection; however, these studies are primarily limited to tendon pathology and muscle strains (14,21,27,36,51,73,75). Although variability exists with respect to exercise selection, progressions, and inclusion time, a majority of programming includes the introduction of mechanical forces designed to work in synergy with respect to PRP. Specifically, most programs focus on initiating early active movement to combat the deleterious effects of immobilization, followed by progressive mechanical loading of the tissues. Progressive mechanical loading (referred to as mechanotherapy) has been shown to produce a favorable anabolic environment and promote tissue regeneration. Understanding the mechanical stimuli to which musculoskeletal cells best respond and the mechanisms these cells use to convert mechanical signals into a cellular response is required to effectively introduce interventions that function synergistically with PRP. A brief, albeit important discussion of the concept of mechanotransduction is necessary to appreciate the benefits associated with mechanical loading. In general, when joints move, mechanical energy is added to an existent equilibrium of the body and stress is introduced to the load-bearing tissues (48). Each muscle, tendon, and bone is composed of cells that are linked together through an extracellular matrix. Cells adhere to these extracellular matrix scaffolds (composed of collagen, glycoproteins, and proteoglycans) through binding of specific receptors of the cell surface (48). Integrins span the cell surface membrane and communicate signals or stress from the external to the internal environment. Essentially, integrins serve as mechanoreceptors given they are the first molecules on the cell surface to sense a mechanical signal (e.g., compression, tension, etc.) and transmit it across the membrane. These signals can alter gene expression, protein synthesis, and metabolism in a manner similar to hormones, growth factors, and cytokines (48). Mechanotransduction is the process by which the tissues undergo an adaptive structural change in response to a mechanical load (53). With mechanotransduction, efforts aim to convert potentially destructive mechanical effects into constructive events that promote recovery through adaptation at the vicinity of the “stressed” cell or neighboring cells (47). Similar to bone getting stronger in response to weightbearing (another form of mechanotransduction), tendon or muscle would remodel in response to loading. Mechanotransduction follows a sequence of 3 steps, which include mechanocoupling, cell-to-cell communication (passing of loading message from one area of tendon to another), and a cellular response such as collagen synthesis or chondrogenesis (53). A necessary requisite for a cellular response is the presence of an appropriate load to the tissue of interest. The physical load that induces mechanotransduction is referred to as mechanocoupling (53). A key point with mechanocoupling is that the “overload” needs to be appropriate, progressive, and short of the point where injury risk presents itself. For example, eccentrically overloading (mechanocoupling) an involved tendon sets off the remaining physical events of mechanotransduction through stimulation of the tendon cell (tenocyte), which in turn leads to cellto-cell communication and a cellular response (collagen synthesis). A detailed discussion of mechanotransduction with illustrations of the cellular processes described can be found in the article by Khan and Scott (53). In summary, for mechanotransduction to occur, a load is applied to the region of interest. This load, through integrins, communicates the load’s message to the cell where the DNA and cytoplasmic elements are located. When the signal of a “load” is received at the cell, a message is communicated to the cell’s nucleus (DNA). Once the nucleus receives the signal, the messenger RNA transcribes the message and shuttles it to the cytoplasm where it is translated into a regenerative response such as collagen synthesis, which would be incorporated into the cellular matrix. The more common exercise principles used to promote mechanotransduction after PRP include active and passive movements, which primarily serve to prevent the deleterious effects of immobilization (30,69,77) and muscle performance activities inclusive of isometric contractions, eccentric overload, and heavy slow resistance training. Fortunately, the mechanical load on the tissues from both isometric and eccentric training induces other positive anabolic responses that further support the effects of PRP in promoting a regenerative environment.


Active and passive motion should be introduced in the early acute phase to deter the negative effects of immobilization. Aside from mitigating movement impairments such as arthrofibrosis, repetitive movements have a favorable effect on cartilage and the intervertebral disc. Unfortunately, much of what we have gleaned in the area of articular cartilage is based on studies investigating the effects of continuous passive movement (CPM) versus immobilization. Nevertheless, these studies provide an understanding of the biophysiological effects that steer early movement and avoidance of prolonged immobilization. Specifically, the biological evidence has shown that integrins in diseased human chondrocytes are sensitive to mechanical stimulation and increase expression of type 2 collagen and proteoglycans (49). In support of these findings, an in vitro study of human arthritic chondrocytes determined that repeated compression influenced chondrogenesis and reduced catabolic events through downregulation of collagenase expression (29). A key point here is that arthritic chondrocytes are characterized by a catabolic phenotype and a repeated compression model downregulated this environment (29). Furthermore, a greater understanding may be gleaned from in vivo laboratory studies of nonhuman knees. In a controlled laboratory study of rabbit knees, the anterior cruciate ligament was resected and the knee joint lines were evaluated 4 weeks after surgery. During the 4 weeks, the rabbits were allocated to CPM, treadmill, or sedentary environments. After 4 weeks, the CPM group had normal articular cartilage, whereas the treadmill and sedentary groups experienced surface abrasion. In addition, the CPM group had lower levels of inflammatory cytokines and a normal level of chondrocytes without damaged collagen fibers when compared with the other groups that had chondrocyte apoptosis, reduced cartilage thickness, and collagen damage (20). Shimizu et al. (77) investigated the repair response of cartilage defects in rabbit knee models and found that joints exposed to CPM had higher chondrocyte numbers than immobilization groups. In addition, delaying CPM for 1 week significantly reduced the beneficial effects on cartilage. Although the aforementioned in vivo studies further highlight the effects of activity and inactivity, one must be cautious in the generalization of these findings to the human clinical environment. Regarding the lumbar intervertebral disc, a body of evidence has suggested that increased diffusion of water content into the disc is associated with pain reduction among individuals with low back pain (8–10). Studies have shown that, when appropriate, posterior to anterior joint mobilization (nonthrust) and prone extension exercises may increase water content within the disc and subsequently be associated with reduced pain (8,10). Assuming that these interventions are appropriate to the individual’s diagnosis, they may be considered as part of routine care.


Although the clinical research on isometric training is limited, evidence does suggest a beneficial response for acute tendinopathy (72). In particular, evidence from one investigation (3-arm [group] trial) has indicated that isometric strengthening produced significant improvements in pain and function as well as reduced tendon thickening (observed for 71% of participants albeit not statistically significant) after intervention (72). Unfortunately, the details of the isometric dosing were unclear, as the authors reported progressing from 3 to 5 times per day with the duration of contraction progressing from 10 to 20 seconds with no mention of repetitions performed with each session. Research on pain ratings during contraction and pressure pain thresholds (sensitivity to pressure) have been favorable regarding isometric training; however, the use of healthy adults in these studies certainly skews the clinical value (46,57). From a dosing perspective, evidence suggests that submaximal isometric contractions are best held until failure, whereas maximal effort contractions may be held for durations as short as 5 seconds (39,43,44,46,57). Regarding anabolic hormones, growth factors, and cytokines, isometric training has been shown to produce increases in growth hormone, testosterone, and IGF (IGF1Ea and MGF) and decreases in myostatin (37,39,43,44). In the study by Hakkinen et al. (39), repeated maximum effort 5-second isometric contractions were used and although favorable changes were noted in all subjects, younger men had a greater response than older men.


A compelling body of evidence exists to support the inclusion of both isometric and eccentric overload training. This evidence is in the form of isolated studies showing a regenerative effect from eccentric training on tendinopathy (23,71) as well as additional studies indicating a favorable anabolic response (hormones and growth factors) and reduced pain when compared with concentric-based training (43,44,54,74). Furthermore, many published PRP studies have incorporated both isometric and eccentric overload training with favorable outcomes (14,21,27,51). Heel drops are a frequently prescribed eccentric exercise for Achilles tendinopathy. A program of twice-daily eccentric overload exercises (Figure 3) resulted in cellular matrix remodeling and reversal of degenerative changes among individuals with Achilles tendinopathy (71). In another investigation of Achilles tendinopathy, the effect of adding PRP to a program of eccentric overload activity was evaluated (27). In the study, both groups showed considerable improvement; however, the addition of a PRP injection offered no additional benefit. In contrast to the aforementioned results, Boesen et al. (14) studied the effect of eccentric training alone versus eccentric with PRP to individuals with Achilles tendinopathy and found that PRP plus eccentric overload has superior outcomes regarding tendon structure and pain reduction when compared with eccentric alone. In another study, heavy slow resistance training offered a comparable benefit (tendon structure and clinical outcomes) to eccentric loading, thus may serve as a valuable alternative or be prescribed concurrently (13). In addition to the clinical benefits and regenerative effects of eccentric overload training on tendinopathy, a considerable body of evidence has identified favorable anabolic growth factor responses. Eccentric overload activity has been shown to induce autocrine and paracrine IGF responses (both IGF-1Ea and mechanogrowth factor [IGF-1Ec]) in tendon and muscle (38,40,43,44), which further supports and influences the desired matrix remodeling effect from PRP. Other benefits of eccentric overload include myostatin inhibition (44). Myostatin is a myokine (cytokine) found in muscle tissue that alters muscle cell differentiation in favor of scar-forming myofibroblasts and has an inhibitory effect on satellite cell activation, hypertrophy, and protein synthesis (15,38). Finally, evidence suggest that repeated bouts of eccentric overload exercise promote a reduction in the inflammatory cytokines (TNF-a and interleukin 8) and an increase in interleukin-10 (anti-inflammatory cytokine) (45). The most appropriate dosage (frequency and repetition range) for eccentric training has yet to be determined (67); however, the protocol that is often followed after a PRP injection was developed by Alfredson et al. (2,14,71). Eccentric overload exercises, using the Alfredson protocol, are performed for 3 sets of 15 repetitions, twice daily, 7 days per week, for 12–24 weeks (2). This protocol has resulted in increased peak torque, decreased pain, and a return to previous level of function in addition to imaging confirmed tendon regeneration (2,71). Recent investigations found that performance of such exercises once daily or 3 times a week may offer similar benefits to twice daily regarding tendinopathy (19,52). Moreover, Stevens et al. (79) found that with respect to clinical outcomes, a “do as tolerated” program revealed comparable clinical outcomes; however, structural changes were not assessed.


Regenerative medicine is a growing field of medicine with rapidly expanding indications and widespread use. Sports medicine and rehabilitation professionals will contend with individuals who have had these procedures. An understanding of the treating physician’s recommendations, basic science, clinical evidence, and postprocedural precautions is necessary for safely reintegrating exercise programming. Although PRP and targeted exercise strategies share common goals of promoting an anabolic or regenerative environment, individuals receiving these procedures are contending with an injury or disorder that requires safety considerations. Specifically, a short period of active rest with the use of ice and NSAIDs is not precluded during the acute postprocedural stage. After the acute period, an ideal microenvironment that promotes appropriately timed movements with progressive overload will seemingly enhance functional outcomes and augment the regenerative response from PRP. Research into postprocedural management is still in its infancy and there is a need for an improved understanding of how mechanical forces may be used to optimize outcomes following PRP procedures.

Download the PDF
Ask the Experts