Abstract
Knee osteoarthritis (OA) causes pain and disability, and autologous adipose-derived stem cell (ASC) therapy has emerged as a regenerative treatment option. This retrospective cohort study compared short-term outcomes of intra-articular ASC injections between patients with moderate (Kellgren–Lawrence [KL] 2/3) and severe (KL 4) OA. Among 242 treated patients, 98 in each group were analyzed after propensity score matching for age, sex, and body mass index. Pain (VAS) and Knee Injury and Osteoarthritis Outcome Score (KOOS) subscales were evaluated at baseline and 1, 3, and 6 months. Both groups showed significant improvements in pain and function, with KL 2/3 patients exhibiting greater gains in KOOS Total, Activities of Daily Living, Sports/Recreation, Quality of Life, and VAS pain compared with KL 4. KOOS pain and symptoms improved similarly in both groups. Analgesic effects increased over time, and no serious adverse events were observed. Mild transient swelling or discomfort occurred in about 5% to 6% of cases. ASC injections provided meaningful symptom relief and functional improvement, particularly in moderate OA, suggesting that preserved joint structure benefits therapeutic efficacy. In advanced OA, benefits were present but attenuated, indicating limited regenerative potential in end-stage disease.
Keywords
Introduction
Knee osteoarthritis (OA) is a chronic, progressive, and degenerative joint disorder that represents one of the leading causes of disability worldwide, particularly among older adults. Epidemiological studies estimate that symptomatic knee OA affects approximately 10% to 15% of adults over 60 years of age, with higher prevalence in women and individuals with elevated body mass index (BMI)1,2. The condition is characterized by degradation of articular cartilage, subchondral bone remodeling, osteophyte formation, synovial inflammation, and changes in joint biomechanics, leading to pain, stiffness, impaired mobility, and reduced quality of life. Patients with knee OA often experience limitations in activities of daily living, recreational and sports activities, and social participation, resulting in significant personal and societal burden 3 .
Radiographic grading of OA severity is widely performed using the Kellgren–Lawrence (KL) classification, which ranges from grade 1 (doubtful narrowing and osteophyte formation) to grade 4 (severe joint space narrowing, large osteophytes, subchondral sclerosis, and deformity) 4 . The KL system provides a standardized framework to guide prognosis, treatment decisions, and research comparability. Clinically, patients with KL 2–3 are considered to have moderate OA, with preserved but compromised cartilage and intermittent functional limitation, whereas KL 4 denotes advanced disease with severe cartilage loss and structural changes, often associated with persistent pain and functional impairment.
Traditional management strategies for knee OA include lifestyle modification, physical therapy, oral analgesics, nonsteroidal anti-inflammatory drugs (NSAIDs), and intra-articular injections of corticosteroids or hyaluronic acid. While these interventions may provide symptomatic relief, they do not address the underlying degenerative process, and efficacy is often transient5,6. Surgical interventions, such as total knee arthroplasty, provide durable symptom relief but are invasive and associated with perioperative risks, particularly in elderly or medically complex patients 7 .
Regenerative medicine has emerged as a potential solution, leveraging the reparative capacity of stem cells to restore joint structure and function. Among these, autologous adipose-derived stem cells (ASCs) are mesenchymal stem cells obtained from subcutaneous fat, which can be harvested with minimal invasiveness. ASCs exhibit multipotent differentiation potential, capable of differentiating into chondrocytes, osteoblasts, and adipocytes, and secrete anti-inflammatory and trophic factors that promote tissue repair, modulate immune responses, and stimulate endogenous progenitor cells8,9. Preclinical studies have demonstrated that intra-articular ASC administration reduces synovial inflammation, preserves cartilage integrity, and improves joint biomechanics in animal OA models10–12.
Clinically, ASC therapy has been associated with improvements in patient-reported outcomes such as the visual analog scale (VAS) for pain, the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), and the Knee Injury and Osteoarthritis Outcome Score (KOOS)13–15. However, the degree to which disease severity affects the therapeutic response remains unclear, while it is reasonable that patients with moderate OA (KL 2/3) may respond more favorably than those with severe OA (KL 4), due to preserved cartilage matrix and a more conducive intra-articular environment for regenerative processes.
This study aimed to evaluate the clinical efficacy and safety of intra-articular autologous ASC injection in patients with knee OA. The primary outcome was a comparison of short-term clinical improvement between patients with moderate OA (KL 2/3) and those with severe OA (KL 4), as assessed by KOOS and VAS for pain. Secondary outcomes included (1) within-group changes in KOOS and VAS for pain to determine whether each KL subgroup experienced significant improvement and (2) assessment of treatment-related complications.
Methods
A retrospective cohort study was performed at a specialized regenerative medicine center, analyzing patients who underwent intra-articular autologous ASC injection for symptomatic knee OA between January 2017 and December 2024. The institutional review board approved the protocol (No. 0010), and all patients provided written informed consent for ASC treatment and the use of their anonymized data for research purposes. Inclusion criteria were adults aged over 18 years with radiographically confirmed tibiofemoral OA (KL 2–4) and persistent pain. Exclusion criteria included isolated patellofemoral OA, KL 1, history of knee surgery, systemic inflammatory arthritis (e.g., rheumatoid arthritis), and any intra-articular injection therapy (corticosteroid, platelet-rich plasma, hyaluronic acid, or other regenerative therapies) administered between ASC injection and the 6-month follow-up. KL grading was independently assessed by two board-certified orthopedic surgeons. In cases of disagreement, the final grade was determined by consensus discussion.
A total of 242 patients were enrolled, and patient demographics, OA type (medial-only or other), BMI, and baseline KOOS and VAS for pain scores were recorded. Patients were divided into the KL 2/3 group (144 patients) and the KL 4 group (98 patients), and propensity score matching was applied to minimize confounding, using age, sex, and BMI as matching variables. A 1:1 nearest-neighbor algorithm without a caliper was employed for propensity score matching, resulting in 98 patients in each group (KL 2: 19, KL 3: 79, KL 4: 98).
Adipose tissue was obtained from the abdominal subcutaneous layer under local anesthesia using a standardized liposuction protocol. Approximately 10 ml of adipose tissue was harvested, processed via enzymatic digestion to isolate stromal vascular fraction, and subsequently cultured to expand ASCs. Adipose-derived cells were processed in a certified cell processing facility according to a standardized protocol. Quality control testing was performed for each lot, including sterility testing, mycoplasma testing, and endotoxin testing, all of which met the predefined release criteria. The final cell product also fulfilled the minimal criteria for mesenchymal stromal cells as defined by the International Society for Cellular Therapy. Cell counts and viability were confirmed prior to injection, ensuring a minimum of 107 viable cells per injection. The ASC suspension was injected into the knee joint under sterile conditions using ultrasound guidance to ensure accurate placement. Patients were monitored for immediate adverse events and instructed to limit high-impact activities for 48 h post-procedure.
Outcomes included VAS for pain and KOOS subscales, Symptoms, Pain, Activities of Daily Living (ADL), Sports/Recreation, QOL, and total KOOS. Assessments were performed at baseline, 1 month, 3 months, and 6 months post-injection. The improvement from the baseline at each time point was calculated and compared between the two groups. Complications and adverse reactions following ASC injections were also recorded.
Data are presented as mean ± standard deviation. For intergroup comparisons, unpaired Student’s t tests were performed to compare VAS for pain and KOOS subscales (Pain, Symptoms, ADL, Sports/Recreation, QOL, and KOOS total score) at baseline, 1 month, 3 months, and 6 months post-injection. Improvement from baseline at each time point was also evaluated. Also, the chi-square test was used to compare the gender ratio, the type of OA lesion (whether it is a medially localized type or not), and rate of reactive pain following ASC injection of the two groups. Statistical analyses were performed using R software (R Foundation for Statistical Computing, Vienna, Austria). The sample size calculation was performed using G*Power 3 (Heinrich Heine Universität Düsseldorf, Germany). Based on our calculations, to perform a Student’s t test with an effect size d of 0.5, a Type I error (α) of 0.05, and a power (1 − β) of 0.95, a minimum sample size of 88 patients per group was required. Statistical significance was considered at P < 0.05.
Results
Baseline characteristics
After propensity score matching, 98 patients were included in each cohort (KL 2/3 and KL 4), with no significant differences in age, sex distribution, BMI, or OA laterality (medial-only vs other patterns) (Table 1). The mean age was 68.6 ± 8.7 years in the KL 2/3 group and 69.2 ± 9.2 years in the KL 4 group. Female participants represented 72.4% in KL 2/3 and 69.4% in KL 4. Mean BMI was 25.6 ± 3.9 kg/m2 and 26.0 ± 4.3 kg/m2, respectively. The proportion of OA types classified as medial-only was 84.7% in both groups. Baseline KOOS scores (Symptoms, Pain, Sports/Recreation, and KOOS total score) were significantly lower in KL 4 patients, reflecting greater disability, while KOOS ADL, QOL score, and VAS for pain showed similar trends but did not reach statistical significance.
Baseline characteristics after propensity score matching.
Note. The values are given as the mean and standard deviation for continuous variables.
KL: Kellgren–Lawrence grade; SD: standard deviation; BMI: body mass index; OA: osteoarthritis; VAS: Visual Analog Scale; KOOS: Knee Injury and Osteoarthritis Outcome Score; ADL: activities of daily living; QOL: quality of life.
P < 0.05.
Clinical outcomes
KOOS subscales
In ADL, Sports/Recreation, QOL, and Total KOOS, improvement was observed in both groups at each follow-up (Fig. 1; Tables 2 and 3). In the KL 2/3 group, the degree of improvement at the 1- and 6-month follow-ups after ASC injection was significantly greater than that in the KL 4 group. In Symptoms and Pain, improvement was observed in both groups at each follow-up while no significant difference in the degree of improvement was observed between the two groups at any follow-up point.
Changes in clinical scores following intra-articular injection of adipose-derived stem cells in the knee (left: group KL 2/3, right: group KL 4): (a) total score, (b) symptoms, (c) pain, (d) ADL, (e) sports/recreation, (f) QOL (a–f: KOOS), and (g) VAS for pain.
Comparison of KL 2/3 and KL 4 groups in clinical outcome improvement by adipose-derived stem cell therapy and the rate of reactive pain after injection.
Note. The values are given as the mean and standard deviation for continuous variables.
KL: Kellgren–Lawrence grade; SD: standard deviation; BL: baseline; VAS: visual analog scale; KOOS: Knee Injury and Osteoarthritis Outcome Score; ADL: activities of daily living; QOL: quality of life.
P < 0.05.
Treatment-Emergent Adverse Events (TEAEs) ≥5% of patients.
VAS for pain
In within-group comparisons, VAS for pain decreased in both groups at 1-, 3-, and 6-month post-ASC injection, indicating clinically meaningful reductions in knee pain. Additionally, a significant reduction in VAS for pain was observed at 6 months after ASC treatment compared with that at 1 month in both groups. At each follow-up point, the KL 2/3 group showed significantly greater improvement than the KL4 group. ASC therapy demonstrated analgesic effects from early stages, with higher analgesic efficacy observed in the KL2/3 group.
Safety and adverse events
Treatment-emergent adverse events occurring in ≥5% of patients are summarized in Table 3. No severe adverse events, infections, allergic reactions, or other complications were observed. All reported events were mild, including transient swelling and localized discomfort (reactive pain), and resolved spontaneously within a few days. No significant differences were observed between the KL 2/3 and KL 4 groups.
Discussion
The present study investigated the short-term efficacy of autologous ASC injections in patients with symptomatic knee OA, stratified by radiographic severity according to the KL classification. By applying propensity score matching, we compared outcomes between moderate (KL 2/3) and severe (KL 4) disease, each with 98 patients. Our current findings can be summarized as follows: (1) improvements were generally greater in KL 2/3 patients, particularly in KOOS total, ADL, Sports/Recreation, QOL, and VAS for pain; (2) both KL 2/3 and KL 4 patients experienced significant improvements in pain and functional outcomes following ASC therapy, although benefits in KL 4 were more limited; (3) in both groups, analgesic benefits became more apparent with time, that is, reductions in VAS for pain continued to occur from 1 to 6 months; and (4) ASC therapy was safe and well tolerated, with no major adverse events reported. Taken together, these results support the therapeutic potential of ASCs for symptomatic knee OA. However, the magnitude of benefit differed according to disease severity: patients with moderate OA achieved more substantial and consistent improvements, whereas those with severe OA experienced more limited gains.
Our findings demonstrate that KL 2/3 patients obtained superior improvements compared to KL 4 patients in several clinically relevant domains, including KOOS total, ADL, sports/recreation, QOL, and VAS for pain. This is consistent with the biological assumption that regenerative therapies are more effective in joints where structural integrity is partially maintained. In KL 2/3 disease, cartilage and subchondral bone are damaged but not completely destroyed, and inflammatory and catabolic activity may be more responsive to modulation by ASC therapy. Conversely, in KL 4 disease, advanced cartilage loss, subchondral sclerosis, and mechanical malalignment may limit the potential for tissue repair, resulting in a smaller treatment effect 16 . Importantly, although KL 4 patients also demonstrated statistically significant gains in all KOOS subscales and VAS scores, the magnitude of these improvements was modest. This indicates that while ASC therapy may offer symptomatic relief in end-stage OA, its clinical efficacy is considerably reduced compared with earlier disease stages. From a clinical perspective, these findings suggest that ASC therapy may be more appropriately considered for patients with moderate OA, where the likelihood of achieving meaningful functional gains appears higher.
An interesting observation in this study was the time course of pain reduction. While both groups demonstrated improvement at 1 month, further reductions in VAS for pain were observed at 6 months. This gradual improvement suggests a cumulative or delayed mechanism of action, consistent with biological modulation rather than transient effect. Unlike corticosteroid injections, which typically achieve rapid but short-lived relief 17 , ASC therapy appears to produce effects that increase over time, possibly due to progressive immunomodulation, paracrine signaling, and recruitment of endogenous repair mechanisms18,19.
Our results align with previous clinical studies demonstrating efficacy of ASCs in knee OA. Jo et al. 20 reported in a randomized controlled trial that intra-articular injection of cultured ASCs improved pain and function in moderate-to-severe OA, with MRI evidence of cartilage repair at higher doses. Other observational studies and meta-analyses have confirmed significant improvements in VAS, WOMAC, and KOOS scores following ASC therapy, with effects sustained up to 12 to 24 months21,22. However, relatively few studies have stratified outcomes by KL grade. Kuwasawa et al. 23 noted that patients with KL 2/3 disease tended to achieve better outcomes than those with KL 4, although improvements were still observed in advanced disease. Our results expand on these findings by applying propensity score matching, thereby strengthening evidence that structural severity influences treatment response. Our findings appear to contrast with those reported by Pers et al. 24 , who observed no significant benefit of a single intra-articular injection of ASC over placebo in patients with symptomatic KL 2/3 knee OA patients, where the sample size was smaller than the current study. The number of cells administered in their study was considerably lower—approximately one-fifth of the cell dose used in our study—potentially insufficient to achieve therapeutic efficacy. Additionally, their trial population was restricted to patients younger than 70 years, which differs from our broader inclusion criteria. Taken together, direct comparison between the two studies is not straightforward.
When considering regenerative options for knee OA, platelet-rich plasma (PRP) and hyaluronic acid (HA) are frequently employed. PRP delivers concentrated autologous growth factors and has shown superiority over normal saline or HA, particularly in early to moderate OA25,26. However, the efficacy of PRP appears to decline in KL 4 patients, where structural damage is advanced 27 . Similarly, HA provides viscoelastic supplementation and modest symptomatic relief, but outcomes are inconsistent and less favorable in severe OA 5 . In comparison, ASCs appear to provide broader benefits across OA severities, including KL 4, with more durable improvements. Unlike PRP or HA, which primarily act through biochemical or mechanical supplementation, ASCs modulate inflammation and stimulate repair processes. Our findings that KL 2/3 patients achieved superior outcomes while KL 4 patients still derived meaningful benefit underscore the potential of ASCs to address both symptomatic and structural components of OA more effectively than other injection therapies.
Some limitations should be acknowledged in this study. First, this was a retrospective study conducted at a single specialized center, which may limit generalizability. This study employed matching based only on sex, age, and BMI. Therefore, the possibility of residual confounding due to unmeasured variables (e.g. occupational status, race/ethnicity, and sports activity level) cannot be fully excluded. Additionally, no sensitivity analyses were performed to assess the robustness of the findings. Second, as the follow-up duration was limited to 6 months, long-term clinical outcome of ASC therapy remains uncertain. Previous studies suggest effects may persist for 12 to 24 months21,22, but extended observation is needed to confirm. Third, MRI or histological assessments were not included, precluding direct evaluation of structural cartilage repair. Fourth, while ASC processing and cell counts were standardized, variability in donor cell characteristics may contribute to heterogeneous responses. Future research should prioritize randomized controlled trials with longer follow-up, including imaging or biomarker endpoints to assess structural modification. Fifth, comparative studies of ASCs versus PRP or HA across KL grades would clarify relative efficacy. Additionally, cost-effectiveness analyses are needed to evaluate the economic viability of widespread ASC therapy in clinical practice.
Conclusion
In summary, our study demonstrates that intra-articular ASC injection provides greater and more consistent improvements in pain and function for patients with moderate knee OA (KL 2/3) compared with those with severe disease (KL 4). Although KL 4 patients also experienced statistically significant improvements across outcome measures, the magnitude of these gains was modest, suggesting a limited therapeutic effect in advanced disease. Pain relief accumulated progressively from 1 to 6 months, consistent with a regenerative mechanism of action. Taken together, these findings indicate that ASC therapy may represent a promising regenerative option for knee OA, with the most favorable outcomes expected in earlier stages and lower OA grades.
Footnotes
Acknowledgements
None.
Ethical Considerations
All procedures were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. All data were fully anonymized, and written informed consent for participation in this study and for publication was obtained from all patients. The name of the IRB: Medical Association Katsujyu-kai Ethics Committee. Ethical approval number: 0010. Date of approval: 8 October 2025.
Author Contributions
N. Nakano: Conceptualization, Formal analysis, Investigation, Methodology, Validation, Writing original draft, Writing review & editing.
M. Tsubosaka: Investigation, Writing review & editing.
T. Kamenaga: Investigation, Writing review & editing.
K. Ishida: Investigation, Writing review & editing.
R. Osaka: Conceptualization, Formal analysis, Methodology.
M. Otsuji: Investigation, Writing review & editing.
N. Nakamura: Supervision, Writing review & editing.
Y. Kuroda: Validation, Writing review & editing.
S. Hayashi: Validation, Writing review & editing.
R. Kuroda: Supervision, Writing review & editing.
T. Matsumoto: Supervision, Writing review & editing.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
Authors can confirm that all relevant data are included in the article.
Statement of Human and Animal Rights
This article does not contain any studies with human or animal subjects.
Statement of Informed Consent
There are no human subjects in this article and informed consent is not applicable.