Intra-Articular Application of Umbilical Cord-Derived Stem Cells in Patients with Chronic Shoulder Pain

Abstract

Background: Chronic shoulder pain is a frequent musculoskeletal complaint that significantly affects function, productivity, and quality of life. Mesenchymal stem cell (MSC)-based therapies have emerged as a potential regenerative option due to their anti-inflammatory and tissue-repair properties. This study aims to evaluate the safety and short-term effectiveness of intra-articular injections of umbilical cord-derived MSCs (UC-MSCs) in patients with chronic shoulder pain. Methods: A retrospective pragmatic observational study was conducted at the Regenerative Medicine Institute in Costa Rica. Medical records were reviewed to extract clinical, sociodemographic, and treatment-related variables. The primary outcome was functional improvement measured with the American Shoulder and Elbow Surgeons (ASES) score. Changes between baseline and the 3-month follow-up were analyzed using paired tests, effect size calculations, and regression models to explore predictors of treatment response. Results: Twenty patients met the inclusion criteria. A significant improvement in shoulder function was observed, with a mean ASES increase of 17.17 points, exceeding the minimum clinically important difference of 12. Sixty percent of patients achieved clinically meaningful improvement. Effect size estimates indicated a large magnitude of change. Regression analyses showed that baseline ASES predicted follow-up scores, while higher UC-MSC doses were associated with greater functional improvement. No adverse events were documented during the study period. Conclusions: The study shows that UC-MSC therapy is a safe, minimally invasive, and clinically beneficial option for chronic shoulder pain. These findings support the therapeutic potential of MSCs and highlight the need for larger controlled studies to validate long-term efficacy and optimize treatment protocols.

1. Introduction

Chronic shoulder pain represents the third most frequent musculoskeletal complaint after low back and knee pain in primary care [1,2]. The most common causes include rotator cuff disease, adhesive capsulitis, calcific tendinitis, impingement syndromes, and degenerative joint changes [3,4]. These injuries have been shown to result in increased demands on healthcare services, reduced work productivity, significant job absenteeism due to illness, and premature retirement or job loss [5,6].

Current treatment regimens are focused on pain management, physical therapy, lifestyle modifications, and addressing the underlying cause of the pain, which frequently requires surgical intervention [7,8,9]. Unfortunately, a significant proportion of patients end up with ineffective results, leading to more pain and movement limitations [10]. These negative results highlight the need for novel regenerative strategies that move beyond symptom control toward restoring tissue integrity.

Mesenchymal stem cell (MSC)-based therapies have received increasing interest as a potential solution. MSCs are multipotent stromal cells capable of differentiating into bone, cartilage, muscle, and adipose tissue while also having immunomodulatory, anti-inflammatory, and paracrine effects that promote tissue repair [11,12]. Preclinical studies in animal models of shoulder injury demonstrate that MSC therapy enhances tendon-to-bone healing, improves fibrocartilage formation, reduces inflammatory cell infiltration, and strengthens biomechanical properties [13,14,15,16].

The amount of clinical information on the safety and efficacy of MSCs has been increasing over the years. Several research groups around the world have focused on demonstrating that this minimally invasive therapy can help patients. Hooper et al. suggest in their review that MSC therapies are safe, demonstrating a reduction in pain and even an increase in joint function in subjects undergoing shoulder intra-articular application [17]. Nejati et al. demonstrated that intratendinous injection of MSCs for full-thickness rotator cuff tears was safe over 36 months, with no local or systemic adverse events, and was associated with sustained improvements in pain and shoulder function, supporting both the safety and potential efficacy of MSC-based therapies in shoulder pathology [18]. Despite the encouraging results, there are still unanswered questions due to the heterogeneity of how MCSs are produced and used.

The present study evaluates the intra-articular administration of umbilical cord-derived mesenchymal stem cells (UC-MSCs) in patients with chronic shoulder pain, aiming to assess their safety and efficacy as a regenerative alternative to conventional management strategies in a non-controlled setting.

2. Materials and Methods

2.1. Study Design, Location and Ethical Approval

This study was designed as a retrospective pragmatic observational study with a pre–post design based on medical records. The objective was to analyze differences in ASES scores among patients diagnosed with chronic shoulder pain who received an intra-articular injection of human UC-MSCs in quantified doses. This analysis was conducted at the Regenerative Medicine Institute (RMI) in San José, Costa Rica, and covered a study period from June 2022 to April 2023. RMI is an outpatient clinic that specializes in umbilical cord-derived mesenchymal stem cell treatments. The clinic operates under the regulation of the Costa Rican Ministry of Health, which has granted RMI the necessary authorization to provide expanded allogeneic mesenchymal stem cell treatments. All data analyzed in this study were derived exclusively from patient information collected within a clinical care setting. This protocol was approved by the ACIB-FUNIN Ethics Committee on January 16, 2024, under approval code CEC-FUNIN-016-2023.

2.2. Population and Sample

A total of 20 medical records were selected using non-probability convenience sampling, with the inclusion criteria being the complete availability of the required data and the representativeness of the conditions of interest. This sample size corresponded to the entire population of patients treated in RMI for shoulder pain between June 2022 and April 2023. Patients aged 18 years or above who had received treatment for chronic shoulder pain at RMI during the specified period were included in the study. Subjects with a history of cancer (active or past), those whose records lacked more than 50% of the requisite information to complete the variable collection sheet, patients with fewer than two records of the clinical scales utilized to assess efficacy, and patients who received treatment outside the designated study period were excluded.

2.3. Study Procedure and Intervention

Prior to any procedure, patients were evaluated using an ultrasound at the affected area. Trained physicians performed a clinical and ultrasonographic diagnosis to plan injection targets. Subsequently, patients were taken to the procedure room, where they underwent light sedation or not, depending on their preference, and were administered UC-MSCs in doses ranging from 7.8 to 38.6 million cells, depending on each patient’s individual needs and ultrasound findings. The mean cell viability at the time of application was 79.4%. The evaluation and treatment protocol is provided in the Supplementary Materials.

2.4. Collection and Processing of UC-MSCs

The collection of UC-MSCs was conducted in accordance with established protocols to ensure the quality and safety of the cells obtained. In accordance with the pertinent regulations, the informed consent of the umbilical cord donors was obtained at the time of term delivery by cesarean section, which included their agreement to the donation of the biological material. A comprehensive assessment of the donor mother’s medical history was conducted to ascertain the absence of hereditary disorders and a strong history of oncological disease. Furthermore, the donor’s health status was monitored by the attending physician prior to delivery. On the day scheduled for the cesarean section, a maternal blood sample was obtained for serological testing, which was analyzed for the presence of transmissible infections. The results of these tests were correlated with the specific umbilical cord collection batch. In the event of a positive result for any infectious pathogen, the umbilical cord was excluded from the donation process. Only umbilical cords that satisfied all established screening and safety criteria were regarded as suitable for utilization.

Immediately following the collection of the umbilical cords, they were transferred to a laminar flow chamber for processing in an aseptic environment. The blood vessels were then removed and washed to remove residual blood. Subsequently, small segments were excised and tested for the presence of Mycoplasma sp. to ascertain that the inoculum was free of this contamination. Following this, the segments were incubated in culture flasks with alpha MEM medium enriched with 10% fetal bovine serum for 15 days. During this period, the medium was replaced every 2–3 days until the cells reached confluence. Once confluence was achieved, the cells were expanded in vitro until passage five. At this juncture, a sample was obtained for quality analysis, including bacterial and fungal culture, re-testing for Mycoplasma sp., and an assessment of multipotentiality through the induction of differentiation into chondrogenic, osteogenic, and adipogenic lineages.

Furthermore, studies examining surface markers were conducted via flow cytometry, with a particular focus on positive markers CD90, CD73, CD105, and CD44. The evaluation of negative markers included CD34, CD19, CD45, and CD11b. Furthermore, genetic stability was confirmed through karyotype analysis. Cells that satisfied all quality criteria were frozen using a designated cryopreservative agent and stored in vials at a temperature of −180 degrees Celsius.

Following thawing, the cells were counted using trypan blue to obtain the total number of cells, as well as the proportion of viable cells prior to each application.

2.5. Data Collection

Data were extracted from digital medical records using a standardized form designed specifically for this study. This form included qualitative variables such as nationality, ethnicity, pathological and non-pathological personal history, and treatments applied, as well as quantitative variables such as age, weight, height, body mass index, number of cells applied, cell viability, and post-treatment results obtained, among others. Data collection was performed by trained personnel to ensure accuracy and uniformity in data collection.

The American Shoulder and Elbow Surgeons Standardized Shoulder Assessment Form (ASES) scales were used to assess the efficacy of the study. The ASES scale is a validated tool for assessing shoulder pain and function, allowing a quantitative assessment of shoulder status in patients with various pathologies [19]. The minimum clinically important differences (MCIDs) used for this study were 12 points, as reported in previous literature [20].

Adverse events were systematically documented by clinicians during the procedure and each follow-up visit based on patient reports and clinical findings. Adverse events were reported in accordance with the Common Terminology Criteria for Adverse Events (CTCAE) version 4 (2017) [21] to ensure standardized reporting and reproducibility.

2.6. Statistical Analysis of Data

Descriptive analyses were conducted, including frequency distributions and measures of central tendency and dispersion, tailored to the nature of each variable. Data visualization was performed using advanced graphical libraries such as ggplot2 and Plotly to ensure clarity and reproducibility.

The distribution of the ASES change score (defined as follow-up minus baseline) was examined prior to conducting hypothesis testing. Normality was assessed using the Shapiro–Wilk test and the Skewness/Kurtosis test and was further evaluated through visual inspection of histograms and Q–Q plots. Both formal tests yielded p-values greater than 0.05, and the graphical evaluations supported the conclusion that the change scores followed an approximately normal distribution.

The difference between baseline and 3-month ASES scores was analyzed using a paired t-test to determine whether treatment produced a significant improvement in shoulder function. To verify that this conclusion did not depend on the assumption of normality, a Wilcoxon signed-rank test was performed as a sensitivity analysis, yielding consistent results. The magnitude of the within-subject effect was quantified using Cohen’s d for paired data, and Hedges’ g was additionally calculated to account for the small sample size.

To further explore factors associated with post-treatment functional scores, an ANCOVA model was constructed with the 3-month ASES score as the dependent variable. The main predictor was the average number of mesenchymal stem cells administered per joint. The model also included baseline ASES, age, and BMI category as covariates, and robust standard errors were applied. Regression coefficients, confidence intervals, and p-values were obtained to assess the contribution of each variable.

A second regression model was used as a sensitivity analysis, in which the dependent variable was the absolute improvement in ASES (the change score). This model included the average UC-MSC dose per joint, age, and body mass index (BMI) category as predictors. Its purpose was to evaluate whether a dose–response relationship became more apparent when analyzing improvements directly rather than the final follow-up values. Because all patients were evaluated at the same fixed time point, no longitudinal or time-varying models were applied.

2.7. Declaration of Large Language Model (LLM) Use

The writing process for this manuscript incorporated the assistance of a large language model (LLM), specifically ChatGPT (OpenAI, Version GPT-5.2). The LLM was used to aid in drafting and refining the manuscript.

All content generated by ChatGPT was thoroughly reviewed, edited, and validated by the authors to ensure its accuracy, coherence, and alignment with the standards of scientific writing. The authors retain full responsibility for the final manuscript and emphasize that the LLM served solely as a tool to enhance the writing process, without replacing intellectual or scientific contributions.

The use of ChatGPT was confined to text generation and did not involve the analysis or interpretation of study data. This declaration is included to ensure transparency and compliance with authorship and accountability standards.

3. Results

3.1. Population Characteristics

A total of 20 patients who received intra-articular injections of UC-MSCs were treated between June 2022 and April 2023. The primary indication for treatment was chronic pain management, which was reported in 100% of cases.

Table 1. Summary of results by clinical and sociodemographic characteristics of the patients treated in the shoulder group.

Characteristics Studied	Variables	Total
		n (%)
Age	Mean	56.7 (SD 11.20)
Nationality	United States	15 (75.0%)
	Costa Rica	2 (10.0%%)
	England	1 (5.0%)
	Brazil	1 (5.0%)
	Spain	1 (5.0%)
Ethnicity	Caucasian	17 (85.0%)
	Latin American	3 (15.0%)
Weight	Mean	79.3 Kg (SD 17.0)
Height	Mean	1.73 m (SD 0.09)
Body mass index	Mean	26.1 kg/m2 (SD 3.7)
Comorbidities	Denies	8 (82.61%)
	Dyslipidemia	5 (25.0%)
	Osteoarthritis	2(10.0%)
	Reumathoid arthritis	2(10.0%)
	Hypertension	2(10.0%)
	Cardiopathy	1 (5.0%)
	Diabetes mellitus type 2	1 (5.0%)
Physical activity	Sedentary	14 (70.0%)
	Recreational	5 (25.0%)
	Athlete	1 (5.0%)
Smoking history	Active smoker	2 (10.0%)
	Non-smoker	18 (90.0%)
Alcohol consumption history	Active alcohol consumer	10 (50.0%)
	Denies having consumed alcohol	10 (50.0%)
History of drug use	Active drug user	2 (10.0%)
	Denies drug use	18 (90.0%)
Previous surgeries	Denies	12 (60.0%)
	Orthopedic surgery	8 (40.0%)

summarizes the clinical and sociodemographic characteristics of the populations who underwent treatment. Figure 1 illustrates the distribution of UC-MSC doses administered per subject

3.2. Measurement of Treatment Outcome

There was a statistically significant improvement in shoulder function following treatment. The paired t-test showed a significant increase in ASES scores at 3 months compared with baseline (p = 0.0004). The median-based Wilcoxon signed-rank test yielded consistent results (p = 0.008), indicating that the improvement remained significant even when not assuming normality. Overall, patients improved an average of 17 points on the ASES scale (Figure 2), reflecting a clinically meaningful change. The effect size analysis showed that the magnitude of this improvement was large, with a Cohen’s d of 0.96 and a Hedges’ g of 0.92 after correction for the small sample.

3.3. Regression Model

An ANCOVA model examining predictors of ASES at 3 months explained approximately 26% of the variance and was borderline significant (p ≈ 0.06). In this model, baseline ASES was the only statistically significant predictor (p = 0.010), indicating that patients with better initial function tended to have higher follow-up scores. The average dose of mesenchymal stem cells per joint showed a positive but non-significant trend (p = 0.113). Age and BMI category were not associated with follow-up ASES scores, as shown in

Table 2. ANCOVA model predicting the change in ASES score at 3 months.

Predictor	β Coefficient	95% CI	p-Value
Average MSC dose per joint	1.37 × 10−6	–3.66 × 10−7–3.10 × 10−6	0.113
Baseline ASES	0.487	0.135–0.840	0.010
Age	0.164	−0.65–0.97	0.672
BMI category	3.18	–7.03–13.39	0.517

.

A secondary regression model using the absolute improvement in ASES as the dependent variable showed a stronger association between MSC dose and clinical response. This model approached overall significance (p = 0.0787) and explained 19% of the variance. The dose per joint variable reached statistical significance (p = 0.039), indicating that higher doses per treated joint were associated with greater improvement in shoulder function. Neither age nor BMI category predicted the magnitude of response, as shown in

Table 3. Linear regression model predicting the ΔASES score at 3 months.

Predictor	β Coefficient	95% CI	p-Value
Average MSC dose per joint	2.28 × 10−6	1.27 × 10−7 to 4.43 × 10−6	0.039
Age	0.239	−0.66–1.13	0.579
BMI category	2.59	10.35–15.53	0.677

. These results suggest a potential dose–response relationship that becomes more evident when analyzing improvement directly rather than follow-up score alone.

3.4. Adverse Events

No adverse events were documented during the 3-month follow-up period. No edema, skin redness, bleeding, hematoma, infection, allergic reaction, ecchymosis, vascular damage, nerve damage, hemarthrosis, or synovitis was observed during or immediately after the procedure during the follow-up period. Furthermore, at the three-month follow-up visit, no long-term complications commonly associated with this procedure were reported, including chronic pain, stiffness, osteomyelitis, septic arthritis, joint deformities, vascular damage, or nerve damage.

4. Discussion

This study assessed the effectiveness and safety of intra-articular UC-MSC injections for treating chronic shoulder pain using a pragmatic retrospective design. The results showed that UC-MSC therapy produced a statistically significant and clinically meaningful improvement in shoulder function. Patients experienced an average increase of 17.17 points in the ASES score three months after treatment. Significantly, 60% of the cohort exceeded the MCID threshold of 12 points, suggesting that the improvements were not only statistically significant but also clinically relevant.

These findings align with prior evidence supporting the therapeutic potential of MSC-based interventions in shoulder pathology. Prospective studies and systematic reviews have demonstrated consistent improvements in pain, range of motion, and patient-reported function following MSC treatment across a variety of shoulder conditions. For example, in the 36-month pilot study by Nejati et al., patients receiving intratendinous autologous adipose-derived MSCs for full-thickness rotator cuff tears demonstrated sustained improvements in pain and shoulder-related disability, as reflected by reductions in the visual analog scale (VAS) and the disabilities of the arm, shoulder and hand (DASH) scores and improvements in the Western Ontario Rotator Cuff (WORC) outcomes over long-term follow-up [18]. Also, Vieira Ferreira et al. described enhanced healing and reduced retear rates when MSCs were used as augmentation during rotator cuff repair [22].

These results suggest that MSC therapies may exert clinically meaningful effects across different shoulder conditions, cell sources, and delivery strategies, with early functional improvement observed in intra-articular approaches and sustained symptom relief reported in longer-term tendon-focused interventions. However, differences in study design, outcome measures, and follow-up duration preclude direct quantitative comparisons and highlight the need for controlled trials to clarify comparative efficacy.

Regression analyses of our dataset provide additional information about potential predictors of treatment response. The baseline ASES score was the only significant predictor of the three-month outcome, suggesting that patients with better pre-treatment function are more likely to maintain superior function post-treatment. However, individuals starting from a lower baseline exhibited larger overall gains, which is consistent with regression to the mean and the greater potential for improvement in more severe cases. Age and BMI were not associated with the magnitude of functional recovery, implying that UC-MSC therapy may benefit diverse demographic groups. Although the ANCOVA model did not detect a clear dose–response relationship when examining follow-up scores, the secondary model using absolute improvement revealed a statistically significant association between higher MSC doses and greater functional gains, as seen in other studies [23]. These results suggest that dose effects may be more easily captured by analyzing change scores directly rather than final status. However, the limited sample size requires cautious interpretation.

The clinical improvements seen in this study are biologically plausible given the known regenerative actions of MSCs. Previous animal and preclinical models have demonstrated that MSCs can downregulate inflammatory mediators, stimulate extracellular matrix remodeling, enhance tendon-to-bone integration, and improve biomechanical properties of injured tissues [16,24,25]. Although imaging biomarkers were not collected in this retrospective analysis, the functional gains observed may reflect these underlying regenerative and immunomodulatory mechanisms.

From a clinical standpoint, these results are relevant for patients who do not respond well to conservative treatment or who are looking for alternatives to corticosteroid injections or surgery. Due to the significant impact of shoulder pain on quality of life, productivity, and long-term disability, minimally invasive regenerative treatments, such as UC-MSC injections, may offer a valuable therapeutic option [5,6]. The absence of adverse events in our study group further supports the safety of this intervention and aligns with the favorable tolerability observed in other MSC studies [23].

This study has several limitations that should be considered. Its retrospective, non-controlled design restricts causal inference and increases susceptibility to selection bias and unmeasured confounding. The small sample size, combined with non-probability convenience sampling, limits statistical power, particularly for identifying predictors of treatment response, and warrants interpretation of the findings as exploratory and hypothesis-generating rather than confirmatory or generalizable to broader populations. Furthermore, the three-month follow-up period precludes assessment of long-term durability or structural change. The lack of imaging outcomes prevents direct correlation between functional improvement and objective tissue repair; therefore, observed gains may reflect symptomatic, neuromodulatory, or biomechanical effects rather than confirmed regeneration. Finally, although cell doses were clinically individualized and exploratory analyses suggested a potential dose–response relationship, dose variability and limited sample size prevent determination of an optimal or standardized therapeutic dose.

Future research should prioritize prospective, randomized, controlled trials with larger samples and standardized dosing regimens. These trials would be essential to distinguish treatment effects from non-specific contextual effects. Incorporating imaging modalities such as MRI or ultrasound, along with biological markers of inflammation or tissue healing, would strengthen mechanistic interpretation. Long-term follow-up is needed to determine the persistence of clinical benefits and their relationship to structural recovery. Additionally, stratified analyses may help identify patient subgroups most likely to benefit from UC-MSC therapy.

In conclusion, intra-articular UC-MSC treatment has been shown to be a safe and promising regenerative option for chronic shoulder pain, yielding meaningful short-term improvements in shoulder function. These findings contribute to the growing body of evidence supporting MSC-based therapies for musculoskeletal disorders and highlight the need for further rigorous investigation to refine clinical protocols and maximize patient outcomes.

5. Conclusions

This study has demonstrated that UC-MSC therapy has significant potential to improve function and alleviate symptoms in chronic shoulder pain subjects, offering improvements in the quality of life and recovery of affected individuals. Despite the limitations, the results obtained are statistically and clinically significant, supporting the use of this treatment in broader patient populations. Future studies should aim to validate these results in larger and more diverse cohorts and explore the interaction of other lifestyle factors to optimize treatment outcomes.