Intraosseous subchondral PRP for knee osteoarthritis: what it is and who it is for

Intraosseous subchondral PRP is platelet-rich plasma placed with image guidance into the bone under the knee joint surface. That means the top of the shin bone (tibial plateau) and the end of the thigh bone (femoral condyle), not just the joint space. It is used for knee osteoarthritis in people trying to hold on to function while they decide about joint replacement. It is almost always given together with a shot into the joint, not by itself. Regen.MD offers it, and we discuss both sides of the evidence with the patient before doing so.

The honest summary is that this is an open question that is moving. A 50-patient randomized trial tested a single added round of bone injections. It found no extra benefit over joint PRP through six months, and more side effects from the shots. A double-blind, placebo-controlled trial published in November 2025 found the opposite. It tested the bone injection as part of a three-shot combination plan. Neither result cancels the other, and both are reported in full below. No orthobiologic has been shown to regrow or regenerate joint cartilage. For advanced knee osteoarthritis, PRP is a bridge therapy prior to arthroplasty.

Key points

• What it is: PRP placed into the bone under the cartilage, at the top of the shin bone and the end of the thigh bone. It uses image guidance and comes with a shot into the joint.
• How it is used: as part of a combined or step-by-step plan, not as a stand-alone shot. Every published clinical study of intraosseous PRP for knee osteoarthritis pairs it with a shot into the joint. No randomized trial has tested it alone.
• The trial that found no added benefit: a randomized trial of 50 adults aged 50 to 65 with Kellgren-Lawrence grade III knee osteoarthritis, at one Indian teaching hospital. Adding two bone injections to a single joint PRP shot gave no significant extra pain relief or better function through six months. The bone-injection group also reported more side effects from the shots.
• The trial that found benefit: a double-blind, multicenter randomized trial of 86 patients with grade III to IV disease in Spain. Both groups got three PRGF shots into the joint. Only the bone component differed: PRGF in one group, saline in the other. The PRGF group improved significantly more in nearly all KOOS and WOMAC areas at 3, 6 and 12 months.
• Who was studied: tightly selected patients. The trials screened out poorly controlled diabetes and high blood pressure, autoimmune disease, high body mass index, daily opioid use, other chronic pain, and serious heart, liver, kidney or blood disease.
• What it is not: a cartilage repair procedure. No orthobiologic has been shown to regrow or regenerate joint cartilage.
• Who it is for: a selected group, picked out by imaging, symptom pattern and metabolic assessment during evaluation. It is not a default option, and not an automatic upgrade to standard PRP.

What is intraosseous (subchondral) PRP?

Joint PRP is placed inside the joint space. Intraosseous or subchondral PRP is placed into the bone just under the cartilage-covered surface. Reaching it accurately takes image guidance and a hollow needle (trocar).

The reason is anatomy. Changes in the bone under the cartilage are part of the disease process in knee osteoarthritis. These include faster bone turnover, hardening, tiny fractures, marrow lesions, and the growth of vessels and nerve fibers. A shot into the joint space does not reach that tissue [10]. Whether reaching it changes patient outcomes is a separate question. It is the question the trials below address.

Does intraosseous PRP work for knee arthritis?

The trial that found no added benefit from a single extra round of injections

A single-blind, parallel-group randomized clinical trial published in Injury in 2022 compared the two approaches head to head [1]. Fifty patients with Kellgren-Lawrence grade III knee osteoarthritis, aged between 50 and 65 years, were randomly assigned. One group got bone plus joint PRP (n = 25). The other got joint PRP alone (n = 25). The trial ran at a single teaching hospital in India. Mean age was 57.12 years in the combined group and 57.00 years in the comparison group.

How the injections were given

The bone-injection group got 18 ml of PRP in total: 5 ml into the tibial plateau, 5 ml into the femoral condyle, and 8 ml into the joint, all under live X-ray. The comparison group got the 8 ml joint shot only. Each patient had one session, not a series, and all 50 were followed for six months [1].

What happened to pain and function in that trial

Both groups improved significantly on pain scores. They also improved on the Knee Injury and Osteoarthritis Outcome Score across pain, symptoms, daily living, sport and recreation, and quality of life. At six months the bone-injection group showed a bigger drop in pain scores. But the difference between groups was not statistically significant (p = 0.422) [1].

On the function scores at six months, the only difference between groups that reached significance was sport and recreation. Pain (p = 0.514), symptoms (p = 0.148), daily living (p = 0.991) and quality of life (p = 0.376) did not differ. A 2026 systematic review in Cartilage reports that the sport and recreation difference favored the joint-only group [9]. The authors’ own conclusion was direct. Bone PRP injections did not give any added benefit over joint PRP until six months, in either pain relief or function [1].

What happened to tolerability in that trial

The trade-off went the other way. Patients in the bone-injection group reported significant side effects from the shots. They also took more doses of acetaminophen than the joint-only group [1]. The registered protocol says the bone injections were done under general anesthesia [2]. That is itself part of the burden being weighed.

That matters for a real decision. Adding two bone injections to a joint shot is a more demanding procedure. In this trial, given once, it bought no measurable extra benefit within the study window.

The trial that isolated the intraosseous component

A forward-looking, double-blind, multicenter randomized controlled trial was published in the Journal of Clinical Medicine in November 2025. It was built to answer a narrower question [4]. Eighty-six patients with Kellgren-Lawrence grade III to IV knee osteoarthritis at two public hospitals in Spain were randomized. One group got plasma rich in growth factors (PRGF) into the bone. The other got a saline placebo into the bone. Both groups then got the same three PRGF shots into the joint. So the only difference between them was what went into the bone.

The bone shot was given under live X-ray: 3 to 4 ml into the tibial plateau and the same amount into the femoral condyle. One to two weeks later came two more joint shots of 6 to 8 ml. Eighty-three patients were analyzed. The main outcome was the KOOS at six months [4].

Both groups improved significantly from baseline at 3, 6 and 12 months. The PRGF bone group improved significantly more than the saline group in nearly all KOOS and WOMAC areas (pain, symptoms, function and quality of life) at all three follow-up points. More patients in the PRGF group reached a meaningful improvement in KOOS pain at every point. But that difference reached significance only at six months (p = 0.0038; p = 0.0606 at three months and p = 0.2267 at twelve). No serious side effects occurred. 15 patients, 7 in the PRGF group and 8 in the saline group, reported pain and inflammation in the first 72 hours that went away on its own [4].

Two cautions belong with that result. Three of the six authors disclose that they are the scientific director of, or scientists at, the company whose PRGF technology was used [4]. And the trial had no third group in which nothing at all went into the bone. Both groups had a hollow needle passed into the bone, a core decompression effect. That mechanical effect cannot be separated from the biology. The authors say so themselves.

The comparative studies that came before

A randomized trial published in Clinical Rheumatology in 2018 studied 86 patients with grade II to III knee osteoarthritis in China. They got one of three: joint plus bone PRP given twice, two weeks apart; joint PRP alone; or joint hyaluronic acid. The combined group had significantly better pain and WOMAC scores, lasting through 18 months [5]. The 2026 Cartilage review judged this trial to be at high overall risk of bias. It found high risk in departures from the planned treatments and in how outcomes were measured [9]. It is a supportive result from a weak design.

An observational study of 60 patients with severe knee osteoarthritis was published in Cartilage in 2019. It matched 30 patients treated with combined bone and joint PRP against 30 treated with joint PRP only. The combined group improved significantly on all KOOS and WOMAC scores at 2, 6 and 12 months. The joint-only group did not improve on any score. A meaningful improvement was reached by 16 of 30 versus 8 of 30 at six months (p = 0.037), and 14 versus 5 at twelve months (p = 0.013) [6]. This was not randomized, and one author discloses a tie to the PRGF maker.

The same Indian group that published the null trial also ran a three-arm randomized trial, published in Injury in 2023. 96 people with grade 3 knee osteoarthritis got joint PRP, combined joint and bone PRP, or joint saline. Both PRP groups beat saline on pain and KOOS at 6 and 12 weeks. Only the combined group showed a significant drop in the bone turnover marker NTX-I at 12 weeks (p = 0.009). It also showed significant drops in extra joint fluid (synovial effusion) and in inflammation of the fluid sac below the kneecap (infrapatellar bursitis) on ultrasound [3]. That is a biological signal at 12 weeks, not a lasting clinical advantage.

The observational cohorts

A look-back study covered 79 patients with grade 3 to 4 knee osteoarthritis at a single Spanish clinic. They got office-based, ultrasound-guided bone and joint PRGF weekly for three sessions. It reported significant improvement in all KOOS areas at a median follow-up of 11 months. 88% of patients had at least a 10-point drop in pain score [7]. The procedure was done with local numbing in an outpatient setting, without sedation.

A real-world group of 86 patients treated with combined bone and joint PRP was published in 2025. It reported a drop in WOMAC starting two months after treatment and lasting to 18 months. Between 55% and 67% of patients met the responder threshold at the various time points [8]. Kellgren-Lawrence grade did not predict response. More white cells in the PRP went with less improvement. That is a preparation factor, not a delivery-route factor.

Both of these lack a control group. Neither can separate treatment effect from natural ups and downs or from expectation. And follow-up data in the real-world group were available for only 42% of the original patients by 18 months [8].

What the reviews conclude, and where they disagree

A 2026 systematic review of shots under the knee cartilage covered 24 studies and 1,109 patients. It found that bone shots may ease symptoms and delay joint replacement in selected patients. It found PRP seems to be a safe option. It rated the overall quality of the studies as modest. Only one of five randomized trials was at low risk of bias [9]. A 2026 full review of intraosseous PRP found the approach improves pain, function and joint structure in moderate-to-severe osteoarthritis, particularly when combined with intra-articular PRP. It also said no randomized trial had tested the bone part alone. Its search closed in May 2025. So it could not have known that gap was about to be filled [10].

On the other side, a 2022 systematic review pooled five studies and 112 patients. It found combined joint and bone PRP linked to significantly better WOMAC scores than joint PRP alone. VAS and KOOS results were mixed. It said the added benefit of the bone route was not well proven [11]. A 2023 meta-analysis pooled 24 PRP randomized trials in 1,344 patients. On pain scores, it found no statistically significant difference between the joint-plus-bone group and the joint-only group [12].

One more thing is worth knowing before you read any summary of this research. The two 2026 reviews describe the same 2022 trial differently. One reports it as showing no difference between groups, except a sport and recreation score favoring the joint group [9]. The other describes the same author’s trials as showing bigger drops in pain and stiffness with the combined plan [10]. When secondary sources split like that, the original trial report is the thing to read.

Who these trials studied, and who they screened out

This is the part of the evidence that rarely reaches a patient. For someone with metabolic disease or several conditions at once, it is the part that decides whether any of the above applies to them.

The exclusion lists, as the studies state them

The 2025 placebo-controlled trial took only patients aged 40 to 75 with a body mass index under 35. It left out poorly controlled high blood pressure or diabetes. It left out autoimmune disease, drugs that suppress the immune system, and active cancer care or follow-up. It left out infection anywhere, arthritis in many joints, a joint bent more than 15 degrees, and an unstable joint. It also left out blood thinners that could not be paused, pregnancy, and positive blood tests for syphilis, hepatitis B, hepatitis C or HIV [4].

The registered plan for the 2022 null trial left out grade IV disease on the Kellgren-Lawrence scale. It left out knee arthritis from inflammation or past infection, and anyone with a past knee replacement or knee fracture. It left out any knee shot in the past three months and corticosteroid (steroid) use in the past four weeks. It left out daily opioid use, backache, and other chronic pain. It left out a body mass index above 40, and blood, heart, liver or kidney disease. And it left out any illness bad enough to stop someone from taking part, including the home exercise part [2]. The registration also lists a wider group than the published report describes: ages 40 to 65 and grades II to III, against the paper’s 50 to 65 and grade III only [1][2].

The 2025 real-world group set out to capture a broader population than trials do. It still left out diabetes mellitus, rheumatoid arthritis, autoimmune disease, clotting disorders, active infection, active tumors, pregnancy and severe heart disease [8]. Its authors note they had no data on body mass index, metabolic syndrome, smoking, or exercise. Any of these could change the result.

For two of the studies cited here, the 2018 randomized trial and the 2019 observational study, the entry and exclusion rules are not reported in the records we could obtain. So no claim is made about who they screened out [5][6].

Why that matters if you have metabolic dysfunction or several conditions

Read those lists together and a pattern shows. The trials kept removing the same people. They removed people with insulin resistance and poorly controlled diabetes. They removed autoimmune and inflammatory disease, obesity over a set line, and heart and kidney disease. They removed people on many drugs, including blood thinners, people on long-term opioids, and people with pain elsewhere. Those are not side details. They describe a large share of the people who actually show up with painful, worn knees.

The logic runs both ways, and both ways have to be stated. A null result in a highly selected group does not prove the treatment fails in the patients who were screened out. It also does not prove it works in them. What it proves is that the question has not been asked. That is the standard gap between how well something works in a trial and how well it works in practice. It is the reason a metabolic assessment is part of evaluation here, not an afterthought.

The trade-off in the least selective studies

There is an uncomfortable symmetry in this research worth naming plainly. The study with almost no medical exclusions at all is the 2023 look-back series. It required only that patients be over 18, with grade 3 to 4 disease and complete outcome data [7]. It is also the study with the weakest design: no randomization, no control group, no blinding. The most rigorous trials are the most selective. The most inclusive studies are the least rigorous. No study in this research is both.

What these designs could not measure

Single-injection trials do not test combination or sequenced care

The 2022 null trial tested one thing: whether two bone injections added to a single joint shot improved six-month outcomes [1]. It did not test a series. It did not test bone PRP spaced out with other injections over weeks. And it did not test it alongside metabolic or inflammation treatment. The 2025 trial that found a benefit gave the bone shot as the opening move of a three-shot plan [4]. The observational groups that report the longest lasting improvement also used series of three, given weekly or every two weeks [7][8]. A trial of a single add-on shot cannot answer whether a step-by-step plan works. And no trial has yet tested bone PRP together with treatment of the metabolic terrain.

Short follow-up does not measure durability

The null trial ended at six months [1]. The 2023 three-arm trial ended at 12 weeks [3]. In the 2025 placebo-controlled trial, the gap between groups was still there at 12 months. But the difference in the share of patients reaching a meaningful pain improvement was significant only at 6 months [4]. Where longer data exist, they come from series without a control group reporting to 18 months [8]. Nobody has published multi-year controlled data on intraosseous PRP for the knee.

Preparation, dose and guidance are not standardized

The studies above used different platelet preparations, volumes, numbers of sessions, imaging, and anesthesia. The 2022 null trial gave 18 ml in one session under live X-ray and general anesthesia [1][2]. The 2025 trial gave 3 to 4 ml per bone site under live X-ray with sedation, then two more joint shots [4]. The 2023 look-back series used ultrasound guidance, 2 ml per bone site, local numbing only, and three weekly sessions [7]. White cell content also varies between preparations. Both a 2023 meta-analysis and a 2025 group study linked leukocyte-poor preparations to better outcomes than leukocyte-rich ones [12][8]. When protocols differ that much, conflicting trial results are what you would expect. Pooling them tells you less than reading them.

Who it is for, and who it is not for

Here, bone-targeted PRP is treated as a selected option, not as an upgrade for everyone already having PRP. We discuss it when imaging and symptom pattern point to the bone under the cartilage as a source. The added burden of the procedure has to be acceptable. And it has to be part of a plan, not a single event.

It is not the right talk when the main pain source is elsewhere, when the imaging does not support the target, or when a less demanding option has not yet been tried. Then a different orthobiologic or a different part of the plan is the better next step. The options are described on our orthobiologics service page.

How risk and benefit are discussed at Regen.MD

The practice position is this. There is still real uncertainty here. The service is offered. And the conversation covers both sides. That means naming the trial that found no added benefit, and the more demanding procedure that came with it. It means naming the newer placebo-controlled evidence and the combination plans with better results. Then it means saying which of those study groups you resemble and which you do not.

When an insurer does not cover something, that is a payer decision, not evidence about whether it works. Patients sign an advance beneficiary notice for non-covered services. Entry is a physician-led Clinical Evaluation with Dr. Gurpreet Singh Padda, MD, MBA, MHP. There, imaging, symptom pattern, metabolic data and past treatment are reviewed together, and a recommendation is made.

“My ethos is to treat all of my patients as I would my own family, with the goal of giving them back their quality of life.” — Dr. Gurpreet Singh Padda, MD, MBA, MHP

Beyond the injection

Knee pain is not only a mechanical problem. Metabolic and inflammatory drivers shape how a joint behaves. As the exclusion lists above show, they are also the exact factors the trials removed. Addressing them is part of the evaluation, not an add-on.

What this is, and what it is not, for advanced knee osteoarthritis

PRP does not regrow or regenerate joint cartilage, and no orthobiologic has been shown to do so. For advanced knee osteoarthritis, PRP is a bridge therapy prior to arthroplasty. It aims at pain and function while the surgery decision is being made. The most hopeful claim the current systematic reviews will make is that bone injections may improve symptoms and delay joint replacement in selected patients [9].

That framing is the honest one. It also means a bridge has to lead somewhere. Regen.MD does provide surgery when conservative care has run out, and Dr. Padda is a surgeon. Surgery follows a completed conservative path. It does not open it.

How long does intraosseous PRP take to work?

The controlled evidence above is reported at six and twelve months. That is a sensible time to recheck, not a promise about your result. Benefit, where it occurs, is judged over months. The plan is revisited based on measured change, not continued by default.

For context on what is used here: Regen.MD makes platelet-rich plasma in two forms. The leukocyte-poor form has no white cells and no red cells. The other form is leukocyte-rich. The form is chosen for the tissue being treated. Both are activated with calcium chloride, at two to three times baseline platelet level, with higher dosing in the six to twelve times range. The target is a measured dose above ten billion platelets per joint.

Frequently asked questions

Sources

1. Barman A, Prakash S, Sahoo J, Mukherjee S, Maiti R, Roy SS. Single intra-articular injection with or without intra-osseous injections of platelet-rich plasma in the treatment of osteoarthritis knee: A single-blind, randomized clinical trial. Injury. 2022;53(3):1247–1253. https://pubmed.ncbi.nlm.nih.gov/35033356/ (Referenced for: 50 patients with Kellgren-Lawrence grade III knee osteoarthritis aged 50 to 65 at a single tertiary care teaching hospital in India; 25 per arm; mean age 57.12 and 57.00 years; 18 ml total in the intraosseous arm comprising 5 ml tibial plateau, 5 ml femoral condyle and 8 ml intra-articular under fluoroscopic guidance, versus 8 ml intra-articular alone; six-month follow-up; significant improvement in both arms; no significant between-group difference in pain relief at six months, p = 0.422; KOOS pain p = 0.514, symptoms p = 0.148, activities of daily living p = 0.991, quality of life p = 0.376; sport and recreation function the only significant functional between-group difference; significant injection-associated adverse events and greater acetaminophen consumption in the intraosseous arm; authors’ conclusion that intraosseous PRP did not provide additional benefit until six months.)
2. Clinical Trials Registry – India, record CTRI/2019/04/018613, Role of Intra-articular and Intra-osseous injections of Platelet Rich Plasma in OA Knee, All India Institute of Medical Sciences Bhubaneswar, via the WHO International Clinical Trials Registry Platform. https://trialsearch.who.int/Trial2.aspx?TrialID=CTRI/2019/04/018613 (Referenced for: registered target sample size of 50 and two-arm design with the same injection volumes as the published trial; registered inclusion criteria of ages 40 to 65, American College of Rheumatology criteria for medial femorotibial knee osteoarthritis, Kellgren-Lawrence grade II/III, pain of at least 20 mm on a 0–100 mm visual analog scale during level walking; registered exclusion criteria of grade IV disease, hemoglobin threshold, inflammatory or post-infectious knee arthritis, prior knee replacement repair, prior knee fracture, any knee injection within three months, daily opioid use, coexisting backache or other chronic pain disorder, body mass index greater than 40, blood, hematological, cardiac, hepatic or renal disorders, oral, topical or intra-articular steroid within four weeks, pregnancy, and comorbidity severe enough to prevent participation in the protocol; intraosseous injections performed under general anesthesia.)
3. Barman A, Bandyopadhyay D, Mohakud S, Sahoo J, Maiti R, Mukherjee S, Prakash S, Roy SS, Viswanath A. Comparison of clinical outcome, cartilage turnover, and inflammatory activity following either intra-articular or a combination of intra-articular with intra-osseous platelet-rich plasma injections in osteoarthritis knee: A randomized, clinical trial. Injury. 2023;54(2):728–737. https://pubmed.ncbi.nlm.nih.gov/36414504/ (Referenced for: 96 persons with Kellgren-Lawrence grade 3 knee osteoarthritis at a tertiary care teaching hospital in India randomized to intra-articular PRP, combined intra-articular and intraosseous PRP, or intra-articular normal saline; outcomes at baseline, 6 and 12 weeks; both PRP groups improved significantly versus saline; significant reduction in NTX-I in the combined group at 12 weeks, p = 0.009; significant reductions in synovial effusion and infrapatellar bursitis in the combined group.)
4. Sánchez Santiuste M, Vaquerizo García V, Pareja Esteban JA, Prado R, Padilla S, Anitua E. Plasma Rich in Growth Factors (PRGF) Versus Saline Intraosseous Infiltrations Combined with Intra-Articular PRGF in Severe Knee Osteoarthritis: A Prospective Double-Blind Multicentric Randomized Controlled Trial with 1-Year Follow-Up. Journal of Clinical Medicine. 2025;14(22):8075. https://pmc.ncbi.nlm.nih.gov/articles/PMC12653810/ (Referenced for: prospective, double-blind, multicenter randomized trial at two public hospitals in Spain; 106 assessed, 86 randomized, 83 analyzed; Kellgren-Lawrence grade III–IV; intraosseous PRGF versus intraosseous saline, both followed by three intra-articular PRGF injections; 3 to 4 ml per bone site under fluoroscopy; KOOS at six months as primary outcome; significantly greater improvement in nearly all KOOS and WOMAC domains at 3, 6 and 12 months in the PRGF arm; minimal clinically important improvement in KOOS pain significant only at six months, p = 0.0038, versus p = 0.0606 at three months and p = 0.2267 at twelve; no serious adverse events, 15 patients with self-resolving pain and inflammation in the first 72 hours; inclusion criteria of ages 40 to 75 and body mass index under 35; full exclusion criteria list; authors’ stated limitations including the absence of a no-infiltration arm and the recommendation to study older and higher-BMI populations; conflict of interest disclosure for three authors.)
5. Su K, Bai Y, Wang J, Zhang H, Liu H, Ma S. Comparison of hyaluronic acid and PRP intra-articular injection with combined intra-articular and intraosseous PRP injections to treat patients with knee osteoarthritis. Clinical Rheumatology. 2018;37(5):1341–1350. https://pubmed.ncbi.nlm.nih.gov/29388085/ (Referenced for: 86 patients with Kellgren-Lawrence grade II–III knee osteoarthritis randomized to combined intra-articular and intraosseous PRP given twice two weeks apart, intra-articular PRP every 14 days, or five weekly intra-articular hyaluronic acid injections; assessment to 18 months; significantly superior visual analog scale and WOMAC scores in the combined group.)
6. Sánchez M, Delgado D, Pompei O, Pérez JC, Sánchez P, Garate A, Bilbao AM, Fiz N, Padilla S. Treating Severe Knee Osteoarthritis with Combination of Intra-Osseous and Intra-Articular Infiltrations of Platelet-Rich Plasma: An Observational Study. Cartilage. 2019;10(2):245–253. https://pubmed.ncbi.nlm.nih.gov/29448817/ (Referenced for: observational study of 60 patients with severe knee osteoarthritis, 30 per group, matched for sex, age, body mass index and radiographic severity, Ahlbäck grade III–IV; significant improvement in all KOOS and WOMAC subscales at 2, 6 and 12 months in the combined group and no improvement in any score in the intra-articular group; minimal clinically important improvement 16 of 30 versus 8 of 30 at six months, p = 0.037, and 14 versus 5 at twelve months, p = 0.013; conflict of interest disclosure.)
7. Ríos Luna A, Fahandezh-Saddi Díaz H, Villanueva Martínez M, Iglesias R, Prado R, Padilla S, Anitua E. Office-Based Intraosseous Infiltrations of PRGF as an Effective Treatment for Knee Osteoarthritis: A Retrospective Observational Clinical Study. Journal of Clinical Medicine. 2023;12(13):4512. https://pmc.ncbi.nlm.nih.gov/articles/PMC10342655/ (Referenced for: retrospective review of 79 patients with Kellgren-Lawrence grade 3–4 knee osteoarthritis at a single Spanish clinic; inclusion criteria of age over 18, grade 3–4 disease, treatment with office-based intraosseous and intra-articular PRGF, completed KOOS before and after, and minimum six-month follow-up, with incomplete data as the only exclusion criterion; 8 ml intra-articular plus 2 ml into each of the tibial plateau and femoral condyle, weekly for three sessions, ultrasound-guided with local anesthesia and without sedation; median follow-up 11 months, interquartile range 7–14; significant improvement in all KOOS domains, p < 0.001; 88% achieving at least a 10-point pain reduction; conflict of interest disclosure.)
8. Catalán JM, Escarrer-Garau G, Estrany-Celià MDM, Parra C, Arbona-González L, Mercader-Barceló J, Dos-Anjos S. Intraosseous and Intra-Articular Platelet-Rich Plasma for Severe Knee Osteoarthritis: A Real-World-Outcomes Initiative. Journal of Clinical Medicine. 2025;14(11):3627. https://pmc.ncbi.nlm.nih.gov/articles/PMC12155101/ (Referenced for: prospective, non-controlled, open real-world study of 86 patients at a private orthopedic clinic in Spain, enrolled August 2020 to August 2024; inclusion of patients over 18 with Kellgren-Lawrence grade 2–4, 88% grade 3 or 4; exclusion of diabetes mellitus, rheumatoid arthritis, coagulopathies, active infections, pregnancy, active tumors, autoimmune diseases and severe cardiovascular disease; three injections at 15-day intervals with the first intraosseous under fluoroscopic guidance; significant WOMAC reduction from two months sustained to 18 months; responders 55%, 60%, 63%, 67% and 58% at 2, 3, 6, 12 and 18 months; follow-up data available for 74%, 64%, 59%, 52% and 42% of patients at those points; Kellgren-Lawrence grade not affecting response; negative correlation between PRP leukocyte count and improvement; authors’ stated absence of data on body mass index, comorbidities, smoking and physical activity; conflict of interest disclosure.)
9. Abu Salem Y, Franceschi C, Conte P, Anzillotti G, D’Arrigo D, Di Matteo B, Kon E. Subchondral Knee Injections: A Comprehensive Systematic Review of Current Clinical Outcomes, Safety, and Arthroplasty Delay. Cartilage. 2026, online ahead of print, published 16 May 2026. https://pubmed.ncbi.nlm.nih.gov/42142000/ (Referenced for: 24 studies and 1,109 patients, mean age 55, mean follow-up 38 months; five randomized controlled trials, only one rated low risk of bias overall; PRP showing a favorable safety profile and lower conversion rates to arthroplasty; conclusion that intraosseous injections may improve symptoms and delay arthroplasty in selected patients; risk-of-bias rating of the 2018 trial as high and of the 2022 trial as some concerns; characterization of the 2022 trial’s sport and recreation difference as favoring the intra-articular arm, p = 0.02; the review’s stated limitation regarding the lack of standardized patient-selection criteria and its effect on generalizability.)
10. García Bordes L, Seijas Vázquez R, Álvarez Díaz P, Castelló Egea A, Cuscó Segarra X, Barastegui Fernández D, Vázquez Gómez M, Ayestaran Calero I, Vizcaíno Navarro S, Laiz Boada P, Ferré-Aniorte A, Velasco Alcalde P, García Balletbó M, Cugat Bertomeu R. Intraosseous platelet-rich plasma for osteoarthritis: a comprehensive review. EFORT Open Reviews. 2026;11(6):512–523. https://pubmed.ncbi.nlm.nih.gov/42227238/ (Referenced for: synthesis of clinical and preclinical evidence from PubMed and Scopus up to May 2025; conclusion that intraosseous PRP, particularly when combined with intra-articular PRP, improves pain, function and joint structure in moderate-to-severe osteoarthritis; the statement that no randomized controlled trial had evaluated intraosseous PRP as an isolated intervention; the subchondral bone pathophysiology rationale; the review’s characterization of the Barman trials.)
11. Patel S, Rajnish RK, Baburaj V, Kumar P, Sharma S, Kumar V. Intraosseous Infiltration of Platelet-Rich Plasma for Knee Osteoarthritis: A Systematic Review of Literature and Limited Meta-analysis. Indian Journal of Orthopaedics. 2022;56(11):1847–1857. https://pubmed.ncbi.nlm.nih.gov/36310547/ (Referenced for: five studies and 112 patients; combined intra-articular and intraosseous PRP associated with significantly better WOMAC scores than intra-articular PRP alone, with VAS and KOOS equivocal; conclusion that the added advantage over intra-articular PRP alone is not well established.)
12. Xiong Y, Gong C, Peng X, Liu X, Su X, Tao X, Li Y, Wen Y, Li W. Efficacy and safety of platelet-rich plasma injections for the treatment of osteoarthritis: a systematic review and meta-analysis of randomized controlled trials. Frontiers in Medicine. 2023;10:1204144. https://pmc.ncbi.nlm.nih.gov/articles/PMC10333515/ (Referenced for: 24 randomized controlled trials and 1,344 patients; subgroup analysis by injection site finding no statistically significant difference between the combined intra-articular plus intraosseous subgroup and the intra-articular-only subgroup on visual analog scale pain, test for subgroup differences P > 0.05; leukocyte-poor PRP more effective than leukocyte-rich PRP for pain.)