PRP vs BMAC for knee osteoarthritis — what is the actual difference?

PRP and BMAC are both autologous biologic injections used in the management of knee osteoarthritis, but they differ substantially in their source material, procedural complexity, cost, and the depth of evidence supporting them.

PRP, or platelet-rich plasma, is produced by drawing the patient's own blood and spinning it in a centrifuge to concentrate platelets to roughly 3 to 5 times their baseline level. The resulting concentrate is then injected into the knee joint. Platelets release growth factors — including PDGF, TGF-β, VEGF, and IGF-1 — that modulate local inflammation and may stimulate tissue repair. The procedure involves only a blood draw, recovery discomfort is minimal, and cost typically falls in the range of $500 to $1,500.

BMAC, or bone marrow aspirate concentrate, requires harvesting bone marrow — most commonly from the iliac crest — which is then concentrated by centrifuge and injected. The biological content is broader than PRP: BMAC contains mesenchymal stem cells, hematopoietic stem cells, platelets, and a wider array of growth factors and cytokines. This richer cellular profile is the basis for its theoretically greater regenerative potential, particularly the capacity to support cartilage formation. The procedure is more invasive, harvest-site soreness typically lasts one to two weeks, and cost generally ranges from $2,000 to $5,000 or more.

On the question of evidence, PRP is the better-studied of the two. Multiple Level I randomized controlled trials have demonstrated PRP to be superior to hyaluronic acid and saline for pain and function in mild-to-moderate knee osteoarthritis, as measured by WOMAC and KOOS scores. The duration of effect is typically 6 to 12 months. There is also evidence suggesting that leukocyte-poor PRP may outperform leukocyte-rich PRP specifically in the osteoarthritis context. Both AAOS and OARSI currently classify the evidence as insufficient to issue a formal recommendation, though support has been growing.

The evidence base for BMAC is less mature. Most published data are Level II to III — cohort studies and case series rather than large randomized trials. Some studies have shown superior cartilage preservation on MRI compared to PRP, and the MSC content provides a theoretical advantage for more advanced disease or structural cartilage defects. However, MSC viability and concentration vary considerably depending on patient age, overall health, and harvest technique. As of 2024, no definitive head-to-head randomized controlled trial has established BMAC as superior to PRP for osteoarthritis outcomes.

Direct comparative trials remain limited. A 2021 study by Centeno et al. suggested BMAC may offer longer durability, but methodological limitations constrain the conclusions that can be drawn. The honest summary of the current literature is that no definitive superiority has been established in high-quality randomized controlled trials.

From a clinical decision standpoint, PRP is a reasonable first-line biologic choice for patients with mild-to-moderate osteoarthritis who are budget-conscious or pursuing their first biologic injection. BMAC warrants consideration in patients with moderate-to-severe osteoarthritis, cartilage defects visible on MRI, and willingness to accept a more invasive procedure in exchange for potential disease modification. In patients over 65 or those with significant comorbidities, PRP is generally preferable because bone marrow cellularity — and therefore BMAC yield — diminishes with age. For patients who have already tried PRP without adequate benefit, BMAC represents a reasonable next step.

Regardless of which option is under consideration, the appropriate sequence begins with orthopedic or sports medicine consultation, weight-bearing X-rays, and MRI to establish osteoarthritis grade and assess cartilage status. A structured physical therapy program running concurrently with any biologic intervention improves outcomes, as does weight optimization where applicable, given its direct effect on joint load.

Knee osteoarthritis disrupts the entire tibiofemoral and patellofemoral arthrokinematic system in ways that should directly inform the choice between PRP and BMAC. The degree of cartilage degradation, synovial dysfunction, and resulting movement impairment are not incidental details — they are the clinical variables that distinguish one intervention from the other.

At the movement level, loss of articular cartilage degrades the roll-glide mechanism of the tibiofemoral joint, so the tibia can no longer roll anteriorly and glide posteriorly during flexion with normal efficiency. Subchondral bone remodeling alters joint congruency, creating asymmetric load distribution — typically the medial compartment bearing 60–70% of load, a pattern that worsens with varus alignment. Synovial inflammation reduces joint lubrication viscosity, increasing friction coefficients during dynamic loading. Mechanoreceptor degradation in the articular cartilage impairs proprioceptive feedback, disrupting neuromuscular control of the joint. Proximally, hip abductors — particularly the gluteus medius — become inhibited through pain-altered motor patterns, increasing dynamic valgus stress and medial compartment loading. Distally, the subtalar joint compensates with excessive pronation to reduce tibial rotation transmission into an already compromised knee. Globally, antalgic gait develops with reduced knee flexion during stance phase, shifting load to the lumbar spine and contralateral limb.

PRP is derived from the patient's own blood through centrifugation, concentrating platelets 3–8 times above baseline. Its mechanism is primarily anti-inflammatory and signaling-based. It delivers growth factors including PDGF, TGF-β, IGF-1, VEGF, and EGF, along with anti-inflammatory cytokines that modulate the synovial environment. It stimulates chondrocyte proliferation and matrix synthesis and reduces catabolic enzyme activity — specifically matrix metalloproteinases — that degrade cartilage. From a biomechanical standpoint, PRP is best suited for early-to-moderate osteoarthritis (Kellgren-Lawrence Grade 1–3). It primarily addresses the inflammatory component of arthrokinematic dysfunction: improving synovial fluid quality leads to better joint lubrication and reduced friction during roll-glide mechanics. Multiple formulations exist, with leukocyte-poor PRP generally preferred for intra-articular use. The evidence for pain reduction and functional improvement at 6–12 months is strong. PRP does not regenerate lost cartilage — it modulates the environment to slow degradation. Effect duration is typically 6–18 months before repeat injection is considered.

BMAC is harvested from the iliac crest, concentrated, and contains a substantially more complex biological payload than PRP. It delivers mesenchymal stem cells — though in lower numbers than are often marketed, typically 1,000–10,000 MSCs per mL — along with hematopoietic stem cells, growth factors comparable to those in PRP plus additional regenerative signaling molecules, and exosomes and extracellular vesicles that drive paracrine regenerative effects. It also carries anti-inflammatory and immunomodulatory properties. BMAC is better suited for moderate-to-severe osteoarthritis (KL Grade 2–4) where structural restoration is the goal. The mesenchymal stem cell component carries theoretical capacity for chondrogenic differentiation — actual cartilage-like tissue formation — though in clinical practice the paracrine signaling effect is likely more impactful than direct differentiation. BMAC addresses both the inflammatory environment and provides a regenerative cellular scaffold. Emerging evidence suggests superiority over PRP for higher-grade osteoarthritis. The procedure is more invasive given the iliac crest harvest, procedural risk is higher, and cost is significantly greater than PRP. If chondrogenesis occurs, the structural benefit may be longer-lasting.

Comparing the two directly: PRP operates through anti-inflammatory and growth factor signaling mechanisms and is best matched to KL Grade 1–3 disease, with minimal or indirect cartilage regeneration, low invasiveness, lower cost, strong evidence, and a repeat interval of every 6–18 months. Its movement impact is primarily through reducing inflammation, which improves range of motion. BMAC operates through regenerative cellular and paracrine mechanisms, is best matched to KL Grade 2–4 disease, carries possible chondrogenic potential, involves moderate invasiveness from the iliac crest harvest, significantly higher cost, moderate-to-strong and growing evidence, and less frequent repeat injection if effective. Its movement impact extends to potential structural restoration and improved arthrokinematics.

Regardless of which biological intervention is chosen, the regenerative environment must be optimized through movement. The following protocol is organized into three phases.

During the first four weeks following injection, the goal is to protect the regenerative environment while maintaining joint mobility. Heel slides for range of motion are performed as 3 sets of 15 repetitions twice daily, targeting 0–120° range, progressing when full range is achieved without pain exceeding 3 out of 10. Quad sets performed isometrically with 10-second holds for 15 repetitions three times daily maintain VMO activation without joint compression. Straight leg raises — 3 sets of 15 daily — provide hip flexor and quad activation without tibiofemoral loading. Ankle pumps, 30 repetitions every waking hour, maintain distal circulation and reduce synovial stasis. Patellar mobilizations in superior, inferior, medial, and lateral directions, held 30 seconds each direction twice daily, maintain patellofemoral arthrokinematics during reduced activity. High-impact loading, deep squatting beyond 90°, and prolonged standing on hard surfaces should be avoided during this phase.

From weeks 4 through 8, the focus shifts to restoring proprioceptive feedback and correcting compensatory patterns. Terminal knee extensions with a resistance band — 3 sets of 20 repetitions daily — restore VMO timing and terminal extension arthrokinematics. Step-ups beginning at 4 inches and progressing to 8 inches, performed as 3 sets of 12 repetitions each leg every other day, provide controlled tibiofemoral loading with an emphasis on eccentric control. Single-leg balance, 3 sets of 30 seconds daily progressing from eyes open to eyes closed to an unstable surface, restores mechanoreceptor-driven proprioception. Clamshells with a resistance band — 3 sets of 15 daily — activate the gluteus medius to reduce medial compartment loading. Lateral band walks, 3 sets of 15 steps each direction three times weekly, build hip abductor strength in a functional position. Progression to Phase 3 is appropriate when single-leg balance reaches 30 seconds eyes closed with less than 2° of knee deviation and pain at or below 2 out of 10 during all Phase 2 exercises.

From weeks 8 through 16, the program addresses the kinetic chain globally and restores functional movement patterns. Goblet squats progressing from bodyweight to loaded, 3 sets of 12 three times weekly, provide full kinetic chain loading with controlled knee tracking; load progresses when depth reaches 90° with symmetric mechanics. Romanian deadlifts — 3 sets of 10 three times weekly — load the posterior chain in a way that reduces anterior tibial shear forces. Lateral step-downs, 3 sets of 10 each leg three times weekly, develop eccentric quad control and valgus correction. Treadmill walking with gait retraining, 20 minutes three times weekly, focuses on symmetric step length, cadence above 100 steps per minute, and elimination of the antalgic pattern. Hip 90/90 mobility work, 2 minutes each side daily, addresses proximal kinetic chain restriction contributing to knee loading asymmetry. Gait normalization benchmarks for this phase include symmetric stance time within 5% side-to-side, knee flexion during loading response of at least 15°, no lateral trunk lean during single-leg stance, and cadence at or above 100 steps per minute at comfortable walking speed.

Phase transitions should not be driven by time alone. Advancing from Phase 1 to Phase 2 requires pain at or below 3 out of 10 at rest, range of motion of at least 0–110°, and no significant effusion — stroke test negative or trace positive. Advancing from Phase 2 to Phase 3 requires single-leg balance of 30 seconds eyes closed, pain at or below 2 out of 10 during step-ups, and a symmetric gait pattern on flat ground. Return to full activity requires a limb symmetry index of at least 85% on the single-leg squat test, pain at or below 1 out of 10 during all functional activities, and no antalgic gait pattern.

The clinical decision framework follows from this analysis. PRP is the appropriate choice when osteoarthritis is early-to-moderate, the primary complaint is pain and stiffness limiting movement quality, and movement assessment shows preserved joint space with primarily inflammatory-driven dysfunction. PRP will reduce the synovial inflammatory load, improve lubrication, and create a better environment for movement restoration. BMAC is the appropriate choice when osteoarthritis is moderate-to-severe, imaging shows significant cartilage loss, and movement assessment reveals structural arthrokinematic compromise — loss of roll-glide mechanics, significant crepitus, joint space narrowing. BMAC's regenerative potential addresses the structural deficit that PRP cannot.

Neither intervention works optimally without a concurrent, progressive movement restoration program. The biological environment created by PRP or BMAC is only as good as the mechanical environment created through corrective exercise and movement re-education. The two must work in concert.

Platelet-rich plasma (PRP) and bone marrow aspirate concentrate (BMAC) are both autologous injectables used for knee osteoarthritis, but they differ meaningfully in their biological mechanisms, evidence base, and cost — and those differences should inform how rehabilitation is structured alongside either treatment.

PRP is derived by centrifuging a patient's own blood to concentrate platelets, typically to 3–5 times baseline concentration. Those platelets release growth factors — PDGF, TGF-β, IGF-1, and VEGF — that modulate inflammation, stimulate synovial cell activity, and may slow cartilage degradation. It functions as a signaling therapy: it instructs existing cells to behave differently. BMAC is drawn from the iliac crest, then centrifuged to concentrate mesenchymal stem cells, hematopoietic progenitor cells, platelets, and growth factors simultaneously. It is both a signaling therapy and a cellular therapy, delivering progenitor cells that carry the theoretical capacity to differentiate into chondrocyte-like cells — though the evidence on true cartilage regeneration remains actively debated.

The clinical evidence favors PRP at this time. Multiple randomized controlled trials and meta-analyses, including Meheux et al. (2016) and Shen et al. (2017), demonstrate statistically significant pain reduction and functional improvement over hyaluronic acid and saline at 6–12 months, particularly in mild-to-moderate osteoarthritis classified as Kellgren-Lawrence Grade 1–3. Leukocyte-poor PRP appears superior to leukocyte-rich formulations specifically for osteoarthritis. BMAC carries more theoretical appeal but less robust clinical evidence. Early studies by Centeno et al. and Shapiro et al. show promising pain and function outcomes, but sample sizes are smaller, protocols vary widely, and head-to-head trials against PRP are limited. The stem cell concentration in BMAC is also highly variable and is often lower than marketed. Critically, neither therapy regenerates significant cartilage in established osteoarthritis. Both are best understood as disease-modifying symptom management tools, not structural cures. The cost differential is also substantial: PRP typically runs $500–$1,500 per injection, while BMAC often costs $3,000–$5,000 or more, with neither typically covered by insurance.

For most knee osteoarthritis patients, PRP is the evidence-supported first choice. BMAC may be considered for higher-grade disease or when PRP has failed, but the cost-to-evidence ratio is less favorable.

Neither injection works optimally without a concurrent strengthening program. This is not a secondary consideration — it is central to the mechanism of benefit. Knee osteoarthritis creates a well-documented neuromuscular cascade. Arthrogenic muscle inhibition (AMI) of the quadriceps is triggered by joint effusion and pain signals via Ib afferent pathways. Even small effusions of 20–30 mL can reduce quadriceps activation by 30–50% through reflex inhibition, independent of pain levels. The vastus medialis oblique (VMO) is disproportionately inhibited, disrupting medial patellar tracking and increasing compressive forces on the medial compartment — the most commonly affected in osteoarthritis. Gluteus medius and maximus weakness follows, producing hip drop and increased knee valgus moment during loading. The result is a self-perpetuating cycle in which inhibition leads to altered loading, which increases cartilage stress, which drives more inflammation, which deepens inhibition. PRP and BMAC reduce the inflammatory signal that sustains AMI, but if neuromuscular control is not simultaneously rebuilt, the structural environment that accelerated the osteoarthritis persists.

Assuming the injection has been performed and the acute inflammatory response has settled — typically 1–2 weeks post-injection — rehabilitation proceeds in two phases.

During weeks 2–4, the focus is neuromuscular activation. Quad sets with biofeedback are performed seated or supine as isometric quadriceps contractions held for 5 seconds, emphasizing VMO activation: 3 sets of 15 repetitions, twice daily. The cue is to push the back of the knee into the surface and feel the inner quad engage. This overcomes AMI without introducing joint compression. Straight leg raises are performed supine, initially without weight: 3 sets of 15, once daily, with the knee locked fully before lifting. Progression to a 1–2 lb ankle weight is appropriate when 15 repetitions feel effortless without quad fatigue. Terminal knee extensions with a resistance band are performed standing, with a light band behind the knee and small-range extension from 30 degrees to full extension: 3 sets of 20, daily. This targets the VMO in a functional position and reinforces the terminal extension motor pattern. Supine hip bridges are performed bilaterally with feet flat: 3 sets of 15, daily, driving through the heels and squeezing the glutes at the top for a 2-second hold. Progression to single-leg bridges is appropriate when bilateral performance is pain-free and symmetric. Side-lying hip abduction is performed at 3 sets of 15 each side, daily, with a slight forward hip tilt and the heel leading rather than the toe. This activates the gluteus medius and reduces valgus collapse during loading.

Progression to phase 2 — loaded functional strength during weeks 4–8 — is appropriate when quadriceps strength reaches at least 70% of the contralateral limb as estimated by handheld dynamometry or single-leg press symmetry, morning swelling is stable, and pain remains at or below 3/10 with phase 1 exercises. The goblet squat to a chair or box is performed with bodyweight initially, to a depth where knee flexion is comfortable — often 60–70 degrees: 3 sets of 12, three times per week, with the chest tall, knees tracking over the second toe, and weight distributed through the full foot. Depth increases progressively as tolerance improves. Forward and lateral step-ups begin on a 4-inch step: 3 sets of 12 in each direction, three times per week, driving through the stepping heel and controlling the descent over 3 seconds eccentrically. Step height progresses to 6 and then 8 inches. Romanian deadlifts are performed with bodyweight or light dumbbells: 3 sets of 10, three times per week. Posterior chain loading through this movement reduces anterior tibial shear and quadriceps compressive demand.

Load progression follows a 10% per week maximum increase when morning swelling is stable. Swelling should be monitored with a tape measure at the joint line; the morning-to-evening differential should remain under 5 mm. If next-day swelling increases more than 5 mm or pain exceeds 4/10 during exercise, load should be reduced by 50% and progression held for one week. Pain during exercise should remain at or below 3/10 as acceptable discomfort; pain above 5/10 is a stop signal.

Return to full activity is supported when quadriceps strength reaches at least 80% of the contralateral limb — with isokinetic testing at 60 degrees per second preferred when available — and when the patient can perform a single-leg squat to 60 degrees without valgus collapse, trunk lean, or compensatory hip hike. For more active patients, a single-leg hop test limb symmetry index of at least 85% is an additional benchmark. Gait should show symmetric cadence, no antalgic pattern, full terminal knee extension in stance phase, and controlled stair descent with pain at or below 2/10 and no lateral trunk shift.

The integrated recommendation is straightforward: PRP is the evidence-supported starting point for most knee osteoarthritis presentations, particularly Grade 1–3. BMAC is worth discussing with an orthopedic specialist for more advanced disease or after PRP has failed to provide adequate relief. Regardless of which injectable is chosen, a structured neuromuscular rehabilitation program should run concurrently. The research consistently shows that biologics combined with exercise outperform biologics alone. The injection modifies the biological environment; the exercise rebuilds the mechanical and neuromuscular environment. Both are necessary.

Both PRP and BMAC demonstrate superior pain and functional outcomes compared to hyaluronic acid injections in knee osteoarthritis. PRP efficacy appears clinically significant and dose-dependent — higher platelet concentrations correlate with better outcomes (PMID 39751394). A landmark randomized controlled trial found that PRP reduced pain but did not prevent cartilage volume loss over 12 months, suggesting symptom relief without structural disease modification in the short term (PMID 34812863). Direct head-to-head evidence shows PRP and BMAC perform similarly, with both outperforming hyaluronic acid, though BMAC may offer theoretical advantages for advanced disease due to its mesenchymal stem cell content (PMID 36913992).

The supporting evidence is drawn from three Grade A sources. Bennell et al., 2021 (JAMA; PMID 34812863) conducted a randomized controlled trial of intra-articular PRP versus placebo in 144 patients with knee osteoarthritis. PRP reduced pain as measured by WOMAC at 12 weeks and 6 months compared to placebo, but did not prevent medial tibial cartilage volume loss on MRI, directly addressing whether PRP modifies disease structure or merely relieves symptoms. Bensa et al., 2025 (The American Journal of Sports Medicine; PMID 39751394) published a meta-analysis of randomized controlled trials demonstrating clinically significant pain improvement with PRP for knee osteoarthritis, with platelet concentration emerging as a key variable — higher concentrations were associated with superior outcomes. Belk et al., 2023 (Arthroscopy; PMID 36913992) conducted a systematic review and meta-analysis comparing PRP, BMAC, and hyaluronic acid injections, finding that both PRP and BMAC showed superior pain and functional outcomes versus hyaluronic acid, with no significant difference reported between PRP and BMAC, and with BMAC's theoretical advantage from mesenchymal stem cell content not definitively translating to superior clinical outcomes in available trials.

Several important evidence gaps and caveats shape how these findings should be interpreted. No large randomized controlled trial has directly compared PRP versus BMAC head to head; the meta-analysis (PMID 36913992) found insufficient direct comparative trials to declare one superior, and most evidence is indirect, with each treatment compared against hyaluronic acid or placebo rather than against each other. The question of structural disease modification remains open — the RESTORE trial (PMID 34812863) showed PRP reduces pain but does not prevent cartilage volume loss at 12 months, and whether longer follow-up beyond 12 months or repeated injections alter this finding is unknown.

Preparation variability is a meaningful practical concern. PRP efficacy depends on preparation method and platelet yield (PMID 39751394), and BMAC similarly varies by harvest technique, patient age, and mesenchymal stem cell viability. Neither treatment is standardized across clinical settings. The evidence base also skews toward mild-to-moderate osteoarthritis, and efficacy in severe disease (Kellgren-Lawrence Grade 4) or in patients older than 70 years is less well characterized.

From a regulatory standpoint, both PRP and BMAC are autologous — derived from the patient's own cells — and are used off-label for osteoarthritis; neither holds FDA approval for this indication, and regulatory oversight is limited. Alignment with current AAOS, AOSSM, or APTA guidelines was not assessed in this search, and current guidance from those organizations should be consulted for formal recommendation status. Cost-effectiveness was not addressed in these studies; BMAC is typically three to five times more expensive than PRP owing to the procedural complexity of bone marrow harvest.