Clinical evidence and critical review of mesenchymal stromal cells for corneal and ocular surface diseases

Abstract

Mesenchymal stromal cells (MSCs), owing to their regenerative and immunomodulatory properties, have emerged as a potential therapeutic option for disorders affecting the cornea and ocular surface. Early-phase clinical studies have begun to demonstrate the safety and, to some extent, efficacy of MSC-based therapies in conditions such as dry eye disease, persistent corneal epithelial defects, ocular chemical injuries, corneal scarring, keratoconus, and limbal stem cell deficiency. However, evidence from some studies suggests that MSC-related improvements may be short-lived. Currently, the appropriate clinical indications, delivery methods, and long-term outcomes remain unclear, necessitating further laboratory and clinical investigations. In this review, we summarize published and ongoing clinical studies on the therapeutic applications of MSCs for ocular surface diseases, including our own group’s experience. We critically evaluate the strengths and limitations of existing studies and highlight gaps and opportunities in this evolving field.

Significance statement.

The integrity of the cornea and ocular surface is essential for functional vision, yet existing treatments fall short in managing severe pathologies affecting these structures. MSCs hold promise as a form of cell-based therapy to address some of these unmet needs. Due to its anatomical accessibility, the ocular surface is especially suitable for localized cell-based interventions. While early studies indicate the safety and potential benefits of MSCs, the suitable clinical indications, optimal delivery, and their efficacy remain unknown. A critical review of existing clinical research can offer valuable insights to further refine and advance MSC-based therapies for corneal and ocular surface disorders.

Introduction

The ocular surface, including the tear film, is critical to the function of the cornea, a transparent and avascular structure essential for vision. Its outermost layer, the corneal epithelium, is vital for preserving corneal transparency and avascularity. This layer undergoes continuous renewal by a population of limbal epithelial stem cells, located at the corneoscleral junction.^1,2 The underlying stroma—which constitutes over 90% of the cornea’s thickness—is composed of a highly organized collagen and glycosaminoglycan–based matrix interspersed with keratocytes, specialized mesenchymal-­type cells.

Mesenchymal stromal cells (MSCs), also known as mesenchymal stem cells, have emerged as a promising therapeutic alternative to conventional treatments such as corneal transplantation.³ These multipotent progenitor cells, derived from various tissues including bone marrow, adipose tissue, umbilical cord, and the cornea, exhibit self-renewal capabilities and multilineage differentiation potential.³ However, as a cell-based therapy, MSCs primarily exert their effects through the secretion of immunomodulatory and trophic factors, which in turn promote endogenous repair mechanisms.³

Current treatments for ocular surface disorders include lubricant eye drops, topical anti-inflammatory agents, ­bandage contact lenses, amniotic membrane (AM) transplantation, and surgical interventions such as corneal transplantation.^4,5 However, these approaches often prove insufficient in chronic, non-healing conditions such as severe injuries and neurotrophic keratitis.⁶ In such cases, MSC-based therapies may offer a novel alternative by leveraging their regenerative properties to modulate inflammation and promote tissue repair.^7,8

To date, 15 clinical trials investigating MSC-based therapies for corneal diseases have been registered, with 7 completed (ClinicalTrials.gov).⁹ However, the optimal source, indications, and dosage/delivery of MSCs remain areas of ongoing investigation. In this narrative review, we critically analyze clinical studies involving human subjects, highlighting their therapeutic potential and remaining knowledge gaps in advancing the use of MSCs in ophthalmic regenerative medicine. Preclinical investigations and single case reports were excluded, and only published clinical studies were included in our analysis (Table 1).

Table 1.

Clinical studies on MSC therapy for corneal and ocular surface diseases.

Study/first author (year)	Møller-Hansen et al.12	Møller-Hansen et al.13	Weng et al.16	Zhang et al.15	Liang et al.19	Funderburgh and Basu et al.25	Tu et al.26	El Zarif and Alió del Barrio et al 36	Calonge et al.29	Bueis et al.30
Clinical trial no. (NCT)	NCT03276358	NCT03276358	Not registered	NCT05784519	NCT02325843	NCT04932629	NCT04736015	NCT02932852	NCT01518867	NCT03173649
PMID	33253910	38049032	23070118	38712345	34333155	_	37896234	33591032	30578758	38152876
Study phase/ status/ design	Phase I/open label	Phase I/II/RCT (double-masked)	Phase I/open label	Phase I/open label (single arm)	Phase I/open label	Phase I/controlled	Phase I/open label (dose-escalation)	Phase I/open label	Phase I/II/RCT (double-masked)	Phase IIa/open label
Indication	DED (SS-related)	DED (SS-related)	DED (GVHD-related)	DED (SS-related and non-SS)	Acute ocular chemical injury	Corneal scarring/ulcers	Persistent corneal epithelial defect	Advanced keratoconus	LSCD	LSCD
MSC Source & Origin	Adipose/allogeneic	Adipose/allogeneic	Bone marrow/allogeneic	Umbilical cord/allogeneic	Bone marrow/allogeneic	Limbal stromal/allogeneic	Bone marrow/allogeneic	Adipose/autologous	Bone marrow/allogeneic	Adipose/autologous
Nature of MSCs	Cryopreserved P2-3	Cryopreserved P2-3	Fresh	Cryopreserved	Not specified (P5/6)	Fresh	Cryopreserved (P4)	Fresh	Fresh Cultured on AM	Cryopreserved → Cultured
Dose and vehicle	2.2–4.4 × 106 in 0.1–0.2 ml CryoStor CS10	4 × 106 in 0.18 mL (CryoStor® CS10)	0.91 × 106 cells/kg IV	5 × 106 in 50 µl (NaCl drops)	5 × 106 in 0.5 mL	0.5 × 106 in fibrin gel	1–6 × 106 in 50–300 µl CryoStor CS5	1–3 × 106 cells ± lamina	250 000 cells	1.6 × 106 (0.4 × 106/quadrant)
Mode of administration	TG lacrimal injection	TG lacrimal injection	IV infusion	Topical (eye drops)	Subconjunctival	Topical embedded in fibrin	Subconjunctival	Intrastromal	Surgical application on hAM	Limbal + corneal + AM patch
Injection site	Lacrimal gland	Lacrimal gland	Systemic	Cornea/conjunctiva (both eyes)	Conjunctiva	Corneal stroma	Conjunctiva	Corneal stroma	Cornea	Limbal + cornea
Sample size	7 eyes (7 pts)	54 pts (20 MSC, 20 veh, 14 obs)	22 pts	16 pts (11 NSDE, 5 SSDE)	16 pts	35 pts (15 in 2018, 20 in 2023)	8 pts (3 dose grps)	13 pts (G1:5, G2:5, G3:3)	28 pts (17 MSC, 11 CLET)	8 pts
Follow-up period	1–16 weeks	1–12 months	1–3 months	1–12 weeks	4–10 weeks	12 months	1–4 weeks	1–36 months	6–12 months	12–86.5 months
Reported outcome	OSDI ↓ 42%, TBUT ↑ 48%, Schirmer ↑ 43%	NIKBUT ↑ 73% at 4w; no sustained benefit vs vehicle	54.5% symptom relief; CD8+CD28¯↑; Th1 ↑, Th2 ↓	Improved SIT, TMH, NIBUT, lipid layer, orifices; better in NSDE; long-term effect in NSDE	81% epithelized; no perf.; all developed LSCD	VA, clarity ↑; NV ↓ vs controls; only 1 surgery required	Safe; 63% defect ↓; no clear dose–response	Safe; G1: VA, thickness ↑; some progression	Comparable to CLET; success 85.7%; safe	75% initial success; only 1 pt w/ sustained benefit at 86.5 months
Adverse events	None	Two with donor-specific HLA antibodies	Not specified	Mild AEs (26.7% discomfort); no serious AEs	None	None	None	None	None	None
Novelty	First-in-human lacrimal MSC delivery	Only randomized DED MSC trial	First systemic MSC use in ocular GVHD	First clinical use of MSC eye drops for DED	Early subconjunctival MSC in burns	Fresh LSSC; fibrin-assisted	First US phase I subconjunctival MSC	Longest follow-up in stromal MSCs	First MSC vs CLET RCT	Longest follow-up in LSCD MSCs
Limitations	No control group	Transient effect, immune sensitization risk	No control group, immune shifts unexplained	Small sample, single-arm, short treatment, no dose comparison	No controls, variable injections	Fresh cells limit scalability	Small sample; no efficacy conclusion	Mixed methods; no control group	hAM and immunosuppression confound	Single center; no control group

Abbreviations: AD-MSC, adipose-derived mesenchymal stromal cell; BM-MSC, bone marrow–derived mesenchymal stromal cell; LSSC, limbal stromal stem cell; MSC, mesenchymal stromal cell.

Cell origin is denoted as Allo, allogeneic or Auto, autologous. Disease categories include ADDE, aqueous-deficient dry eye; DED, dry eye disease; GVHD, graft-versus-host disease; LSCD, limbal stem cell deficiency; NSDE, non-Sjögren’s dry eye; PCED, persistent corneal epithelial defect; SS, Sjögren’s syndrome; SSDE, Sjögren’s syndrome dry eye. Treatment comparisons reference CLET, cultivated limbal epithelial transplantation. Cell carriers include AM, amniotic membrane. Administration routes are abbreviated as IV, intravenous; TG, transconjunctival. Clinical outcomes are summarized with NIBUT, non-invasive break-up time; NIKBUT, non-invasive keratographic break-up time; OSDI, Ocular Surface Disease Index; SIT, Schirmer I test; TBUT, tear break-up time; TMH, tear meniscus height. Additional abbreviations include AEs, adverse events; diff, difference; grps, groups; NV, neovascularization; pts, patients; perf., perforation; VA, visual acuity.

Dry eye disease

Dry eye disease (DED) affects up to 50% of the population, with a higher prevalence in older adults, females, and individuals with autoimmune disorders.¹⁰ It can also be secondary to chronic inflammatory conditions such as graft-versus-host disease (GVHD) and Sjögren’s syndrome (SS).¹¹

Møller-Hansen et al. investigated the safety of allogeneic adipose–derived MSCs (AD-MSCs) injection into the lacrimal gland for severe aqueous-deficient dry eye (ADDE) associated with SS.¹² In this open-label phase I trial, investigators performed a single transconjunctival injection of cryopreserved allogeneic AD-MSCs into the lacrimal glands of one eye of 7 patients (mean age: 59 years). The administered dose was 2.2 × 10⁶ or 4.4 × 10⁶ MSCs per lacrimal gland in 0.1 mL (n = 4) or 0.2 mL (n = 3). Over a mean follow-up period of 4 months, no adverse events related to the treatment were observed. Additionally, after 4 months, a significant decrease in dry eye symptoms (42%) was reported based on the Ocular Surface Disease Index (OSDI), along with enhanced lacrimal gland function, as indicated by increased tear break-up time (48%), Schirmer’s I test scores (43%), and reduced tear osmolarity (7%).¹² While this first clinical trial on AD-MSC injections for SS-related ADDE provided a favorable safety profile, its clinical efficacy and broader applicability could not be fully assessed given the lack of a control group. A double-masked, randomized trial by the same group was conducted to evaluate the efficacy of cryopreserved allogeneic AD-MSCs in 54 patients with severe ADDE due to SS over a 12-month follow-up.¹³ Participants received a single transcutaneous injection of either AD-MSCs (0.18 mL, 4 × 10⁶ MSCs per lacrimal gland; n = 20), vehicle (the cryopreservative CryoStor^® CS10; n = 20) into the lacrimal gland of one eye, or were assigned to an observational group receiving only lubricating eye drops (n = 14). In nearly all of the outcome measures, including both symptom scores and clinical tests, both the AD-MSC and vehicle groups showed improvement compared to the observation group; however, there was no difference between the AD-MSC and vehicle groups. The only notable difference was that the AD-MSC-treated eyes showed a significant increase (73%) in non-invasive keratographic break-up time (a measure of tear film stability) at 4 weeks compared to the vehicle group; however, this effect was transient, peaking at 4 weeks and was not detectable at subsequent time points. Furthermore, no significant differences were observed between groups in fluorescein staining, tear osmolarity, or Schirmer test scores. This suggests that either MSCs only had a small impact beyond the vehicle or that the vehicle itself had a therapeutic or placebo effect. Interestingly, two patients developed new HLA class I donor-specific antibodies—one at four months and the other at 12 months post-transplant. The emergence of donor-specific antibodies raises questions about long-term immunogenic risks and potential implications for repeat MSC treatments. The observed transient effects of the MSCs in the treatment group, may also suggest limited survival of the donor MSC. Achieving long-term MSC engraftment has been a challenge in the field, either due to immunological clearance (in the case of allogeneic cells) or poor MSC survival.¹⁴ Overall, the study by this group remains the only one to date that specifically evaluated the efficacy of MSC therapy, while also offering incidental insights into potential immunogenic responses.¹³

The first-in-human clinical trial using umbilical cord–derived MSCs eye drops for refractory dry eye disease (DED), including both non-Sjögren’s dry eye (NSDE) and Sjögren’s syndrome–associated dry eye (SSDE) subtypes, was conducted by Zhang et al.¹⁵ in a prospective, open-label, single-arm design. Sixteen patients (11 NSDE, 5 SSDE) received bilateral instillation of MSC eye drops (5 × 10⁵ cells in 50 μL per eye) twice daily for 2 weeks. In the NSDE group, significant improvements were observed, including a 13.4-mm increase in Schirmer I test (SIT), a 0.123-mm increase in tear meniscus height (TMH), and a 4.42-s increase in average NIBUT. Subjective symptoms also improved, with reductions in OSDI scores noted at both 2 weeks and 12 months. The SSDE group showed more modest benefits, with a 5.5-mm increase in SIT and a 0.072-mm rise in TMH, while tear stability showed minimal sustained improvement. Tear cytokine profiling demonstrated reductions in IL-6 and IL-17A and elevated Mucin 5AC, indicating anti-inflammatory and mucin-enhancing effects. No serious adverse events were reported. Overall, this noninvasive MSC delivery method showed encouraging safety and efficacy in NSDE, although the therapeutic effects in SSDE were comparatively limited and less durable.

In another study, Weng et al. explored the potential of systemic delivery of MSCs for refractory DED in chronic GVHD.¹⁶ In this study, 22 patients with moderate to severe ocular GVHD received intravenous allogeneic bone marrow–derived MSCs (BM-MSCs) at a dose of 0.91 × 10⁶ cells kg⁻¹, with additional doses administered based on response at 2 weeks. All had refractory chronic GVHD with persistent dry eye despite appropriate immunosuppressive treatment (systemic and topical administration of corticosteroids or cyclosporine A). Based on the outcomes, the investigators divided their study patients into responders and nonresponders and analyzed them separately. At 3 months, 12 patients (54.5%), classified as the effective group, demonstrated significant symptomatic improvement based on NIH consensus criteria, including enhancements in OSDI scores and Schirmer test results, whereas 10 patients (45.5%), categorized as the ineffective group, showed no meaningful clinical response. Flow cytometry analysis of peripheral blood mononuclear cells revealed a significant increase in CD8⁺CD28⁻ T cells (regulatory T cells) in the effective group, while CD4⁺CD25⁺ T cell levels remained unchanged. In the same group, plasma levels of Th1 cytokines—including IL-2 and interferon-γ—increased by 70.5%, whereas Th2 cytokines (IL-10 and IL-4) decreased by 5.5%. No significant immunological changes in plasma were observed in the ineffective group.¹⁶ This study was the first to evaluate systemic MSC infusion for GVHD-associated DED with concomitant T cell changes in some patients. However, it seems counterintuitive that they found MSC treatment to increase Th1 cytokines in the effective group, given that Th1 signaling is generally associated with more pathologic changes in ocular GVHD.¹⁶ This paradoxical finding may reflect assay limitations or statistical variation due to the small sample size.¹⁶ It remains unclear whether the observed improvements in signs, symptoms, and cytokine profiles in the effective group were attributable to MSCs or mostly a reflection of the natural fluctuations of the disease, which can wax and wane.¹⁶ Additionally, the long-term durability of these effects was not assessed due to the lack of extended follow-up. Overall, the small sample size, lack of a placebo control, and short follow-up limit the conclusions that can be drawn about efficacy.

Ocular surface chemical injury

Corneal chemical injuries lead to inflammation, neovascularization, and fibrosis, severely affecting vision.¹⁷ Current acute treatments, such as steroids or AM, offer some benefits, but long-term outcomes remain limited.⁴ MSCs have demonstrated potential in promoting corneal healing by enhancing epithelial proliferation, reducing apoptosis, and modulating inflammation in both in vitro and in vivo experimental models.^7,18

In an open-label, single-arm study, Liang et al. evaluated the safety of subconjunctival, culture-expanded, allogeneic BM-MSC injections (5 × 10⁶ cells 0.5 mL⁻¹ using passage 5 or 6 cells) following AM transplantation in 16 patients with acute severe ocular burns (Dua’s grade IV–VI; mean age, 43 years) within 2 weeks of injury.¹⁹ Eight eyes received 1 injection, 6 had 2, and 2 received 3 because of persistent epithelial defects. The primary outcome was the rate of complete corneal epithelialization without perforation. At 24 weeks, 13 patients (81%) achieved complete epithelialization, 5 (31%) showed improved visual acuity, and 5 (31%) had minimal symblepharon severity. No perforation, necrosis, or infection occurred, and no eyes required emergency keratoplasty. While the results are promising, the study is limited by the lack of a control group and variability in injection frequency.⁷ The slitlamp imaging data indicate that the treatment was insufficient for ocular surface regeneration, as all patients developed partial or total limbal stem cell deficiency (LSCD) by the final visit. Perhaps the absence of perforation, a common complication in very severe injuries, could potentially be attributed to MSC injections. Overall, while the study offers valuable insights in terms of the safety of this approach, it highlights the possible limitations of subconjunctival MSC injection, as it did not substantially prevent fibrosis and vascularization of the cornea.

Persistent corneal epithelial defects and corneal scarring

Severe corneal epithelial disease, characterized by recurrent epithelial breakdown or persistent corneal epithelial defects (PCEDs), is frequently associated with LSCD, neurotrophic keratitis, and other severe ocular surface diseases.^20-22 PCED, resistant to standard care, remains a risk for stromal degradation and corneal perforation; in such cases, MSC-based therapies may offer a promising alternative for promoting epithelial healing.²³

Basu et al. evaluated the long-term safety and efficacy of a minimally invasive approach for delivering allogeneic limbal stromal stem cells (LSSC—a type of MSC) to enhance corneal transparency and improve vision in severe corneal disease.^24,25 Limbal MSCs were isolated from 1-clock-hour limbal biopsies of cadaveric corneoscleral rims and expanded in culture. In a pilot randomized, controlled trial,²⁵ they investigated the effects of fresh limbal MSCs in patients with anterior corneal pathologies. For the delivery, the MSCs (0.5 × 10⁶) were mixed with fibrinogen, applied to the debrided corneal stroma, activated with thrombin to form an adherent fibrin gel, and then covered with a bandage contact lens.²⁵ Patients in the control group received fibrin gels without MSCs along with standard therapy. The group also reported a variation of their technique, whereby the MSCs were first encapsulated in alginate hydrogels prior to delivery in fibrin gels.²⁴ The study included eyes with acute burns, nonhealing ulcers, and postkeratitis scars. Across multiple follow-ups—with longer-term data,^24,25 MSC-treated eyes showed superior visual acuity, corneal clarity, and reduced vascularization compared with controls. While the variability in underlying pathologies may have influenced treatment responses, this study was one of the first to examine the potential use of corneal-derived MSCs for corneal scarring and anterior corneal pathology, with very promising results. Notably, the use of fresh cells (not cryopreserved) likely enhanced engraftment.²⁴

In the first US-based phase 1b clinical trial, our team evaluated the safety and tolerability of subconjunctival allogeneic BM-MSCs (Passage 4 cryopreserved in Cryostor CS5 media) in patients with PCEDs secondary to neurotrophic keratitis or LSCD.²⁶ This dose-escalation study included 3 groups: the first group received 1 × 10⁶ MSCs/50 μL, the second 3 × 10⁶ MSCs/150 μL, and the third 6 × 10⁶ MSCs/150 μL, administered as two separate injections. Safety was assessed at day 28, with ocular and systemic adverse events monitored for 90 days. All 8 participants (100%) completed the study, and MSC treatment was well tolerated at all doses, with no ocular or systemic toxicity. A reduction in defect size was observed in 5 patients (63%), though no clear dose–response relationship emerged. The variability in the timing of epithelial healing following MSC administration makes it difficult to establish a direct causal relationship. The study, however, provided valuable data on the safety and tolerability of escalating MSC doses across different concentrations.²⁶ A phase 2 trial is currently underway to further evaluate the efficacy of MSCs in severe corneal epithelial disease.

Limbal stem cell deficiency

LSCD, often resulting from chemical injury, severe inflammation, long-term contact lens wear, or congenital disorders, leads to impaired corneal healing and vision loss.^27,28

In a randomized, double-masked phase I and II clinical trial, Calonge et al.²⁹ compared cryopreserved allogeneic BM-MSCs (n = 17, 20–200 × 10⁶ cells mL⁻¹, ∼250 000 cells) with allogeneic cultivated limbal epithelial transplantation (CLET; ­n = 11) for treating severe bilateral limbal LSCD. Cells designated for CLET or BM-MSC were cultured on de-epithelialized human amniotic membrane (hAM) and harvested upon reaching ∼90% confluence (∼250 000 cells). After confirming the cultures were pathogen-free, they were delivered as fresh sheets. A total of 23 patients (82%) completed the 12-month follow-up, with no reported intraoperative or postoperative complications.²⁹ The study demonstrated that BM-MSC treatments were as safe and effective as CLET, with global success rates at 6 and 12 months of 72.7% and 77.8% for CLET and 76.5% and 85.7% for MSCs, respectively, showing no significant difference between the 2 groups. Kaplan–Meier analysis of transplant survival showed no statistically significant difference between groups at 6 and 12 months. At 12 months, the central corneal epithelial phenotype improved in 71.4% of MSC cases and 66.7% of CLET cases, with no significant difference between the groups. One confounding factor in the study was the use of hAM as a scaffold, which prevents discerning the independent effects of MSCs, as, given hAM’s intrinsic regenerative properties, the observed benefits may not be solely attributable to MSCs. Additionally, systemic ­immunosuppression—including oral mycophenolate mofetil, cyclosporine A, or ­azathioprine—was administered for 12 months in both groups, raising uncertainty about the effects of these medications independent of the cells (MSCs or cultivated epithelial cells).

Bueis et al. evaluated the safety and feasibility of autologous AD-MSCs in LSCD through a phase IIa trial.³⁰ Cryopreserved AD-MSCs were thawed and cultured for 1 week before transplantation. Following the removal of 6 mm of central corneal epithelium, the patients (n = 8) received 0.4 × 10⁶ autologous AD-MSCs injections per limboconjunctival quadrant, totaling 1.6 × 10⁶ AD-MSCs per eye in the limbal area. An additional 0.4 × 10⁶ cells were applied over the cornea for 20 min, followed by coverage with an AM patch. At 1 year, no adverse events were reported, 5 patients (62.5%) showed improvements in uncorrected visual acuity, and 6 (75%) had successful transplants based on a composite criterion, including absence of pain, improvement in visual acuity, absence of corneal epithelial ulcers, and reduction in corneal neovascularization, as assessed through slitlamp examination and fluorescein staining. Benefits persisted in 4 patients (50%) at 24 months, and by the final follow-up at 86.5 months, no epithelial defects were observed. However, only 1 patient (12.5%) maintained improvements across all measures.³⁰ One notable strength of this study is its meticulous cell preparation process, which involved cryopreservation followed by thawing and culturing a week before transplantation, ensuring optimal cell viability and consistency. However, the study also had certain limitations. It did not include supporting images, and key ass­essments, such as staining grades, were not thoroughly documented. Additionally, each patient received treatment based on their individual clinical course, making it challenging to determine whether the observed effects were directly attributable to MSC therapy. The lack of a control group further limited the ability to draw definitive conclusions, and the therapeutic benefits appeared to diminish over time, with only 1 patient maintaining improvements at the final follow-up. The progressive decline in efficacy observed in long-term follow­up—such as the 12.5% sustained benefit at 86.5 months in the Bueis et al.³⁰ study—suggests that repeat MSC administrations or strategies for prolonged delivery may be necessary to maintain therapeutic effects. In our opinion, repeat MSC injections may yield attenuated clinical responses compared with the initial dose, potentially due to heightened immunologic reactions following sensitization to allogeneic MSCs after the first exposure. This effect could be partially mitigated by using MSCs from a different donor for subsequent doses to reduce immune recognition. Future studies should explore optimal dosing intervals and delivery platforms to enhance durability.

Keratoconus

Keratoconus is a progressive eye condition that affects both eyes asymmetrically, leading to gradual thinning and steepening of the cornea, irregular astigmatism, and a noticeable decline in visual acuity.^31,32 Treatments such as rigid gas-permeable contact lenses, corneal cross-linking, and corneal transplantation aim to preserve or restore vision and slow progression but may not be suitable for all patients.³²

El Zarif and Alió del Barrio et al. conducted a series of clinical trials from 2017 to 2021 to evaluate the safety and efficacy of stromal augmentation using MSCs and corneal laminas in patients with advanced keratoconus.^33–36 One study³⁶ involved 13 patients divided into 3 groups: Group 1 (G1) received freshly expanded autologous adipose-derived MSCs (3 × 10⁶ cells mL⁻¹, n = 5) delivered into a femtosecond laser–created stromal pocket; Group 2 (G2) received a 120-μm decellularized corneal stromal lamina (n = 5); and Group 3 (G3) received a 120-μm lamina recellularized with autologous AD-MSCs (1 × 10⁶ cells mL⁻¹, n = 3). No treatment-related complications were observed in any group, confirming long-term safety. Group 1 showed significant improvements in uncorrected and corrected visual acuity, along with moderate increases in corneal thickness and volume, while refractive values remained stable. Moreover, the comparison between AD-MSC–recellularized lamina and decellularized stroma alone revealed no significant advantage for the MSC-treated group, raising questions about the added value of MSCs in this therapeutic strategy.

Although this study represents one of the longest follow-ups to date evaluating the safety of intrastromal MSC therapy in advanced keratoconus, the outcomes were heterogeneous. Interpretation was further complicated by the varying treatment strategies—only 1 group received MSCs alone, while others underwent additional lamina implantation. This variation made cross-group comparisons challenging. While the trials confirmed the safety of AD-MSC interventions, their broader clinical utility remains uncertain. Despite modest improvements in corneal thickness and visual function, keratometric indices showed minimal change, and in some cases, mild progression of keratoconus was still noted. These findings suggest that AD-MSC therapy alone may not be sufficient to halt ectatic progression in advanced disease.

Future directions

Despite encouraging early results, several challenges remain before MSC-based therapies can be widely adopted in clinical ophthalmology. These include variability in MSC sources, donor characteristics, expansion protocols, cryopreservation, and delivery—all of which limit reproducibility across studies. Post-thaw viability loss and inconsistent dosing strategies further complicate clinical translation.

Topical application faces hurdles such as limited ocular retention and immune-mediated clearance, often resulting in short-lived effects. This highlights the need for improved delivery platforms—such as hydrogels, encapsulation systems, or MSC-derived secretomes—that may enhance survival and efficacy while reducing immunogenicity. In addition, long-term data suggest that the therapeutic effects of MSCs may diminish over time, indicating a potential need for repeat administration or sustained-release strategies to maintain efficacy.

Future efforts should focus on developing consensus guidelines for MSC preparation and administration, alongside multicenter randomized trials with consistent endpoints. Such strategies will be essential to optimize safety, efficacy, and regulatory approval, ultimately advancing MSC-based treatments for corneal and ocular surface diseases

Conclusion

The cornea offers a unique opportunity for localized cell-based therapies due to its easy accessibility, which makes both the application of the treatment and the assessment of the therapeutic outcomes easier. The early-phase clinical studies reviewed here all demonstrate an excellent safety profile for MSCs administered locally to the ocular surface. However, their efficacy has been less consistent, partly due to the lack of control groups; variability in cell-based treatment doses and the participants’ ocular surface conditions, delivery methods, sources of MSCs; administration of concomitant therapies; and small sample sizes of the studies. Findings from some clinical trials suggest that the therapeutic benefits achieved with MSC treatments were not maintained over long-term follow-up.

In general, we believe that MSC therapy in the cornea may be better suited for acute or nonprogressive conditions—such as injuries or scarring—where their transient paracrine effects could provide therapeutic benefits by potentially preventing complications or reducing the need for more advanced interventions. In contrast, their effects appear less durable in chronic diseases such as DED, ocular GVHD, or LSCD, where long-term immunomodulation may be required for sustainable benefit.

The application of MSCs in keratoconus may potentially offer a distinct therapeutic paradigm, focused on structural stromal regeneration via direct intrastromal transplantation rather than immune modulation, a key mechanism of MSCs’ therapeutic effects. Large, multicenter randomized trials are essential to validate and refine different treatment strategies utilizing MSCs for corneal disease. Finally, considering the accessibility of the ocular surface, MSC-derived secreted and extracellular vehicles (EVs) factors may offer an effective, cell-free therapeutic alternative—such as topical eye drops—potentially bypassing barriers related to cell viability and immune response.

Author contributions

Grace C. Tu (Writing—original draft [equal]), Seyyedehfatemeh Ghalibafan (Methodology [equal], Visualization [equal], Writing—original draft [equal], Writing—review & editing [equal]), Farshad Abedi (Writing—review & editing [equal], Charlotte E. Joslin (Writing—review & editing [equal]), Reza Dana (Writing—review & editing [equal], Peiman Hematti (Writing—review & editing [equal]), Ali Djalilian (Conceptualization [equal], Data curation [equal], Funding acquisition [equal], Investigation [equal], Methodology [equal], Project administration [equal], Supervision [equal], Validation [equal], Writing—review & editing [equal])

Funding

This work was supported by R01 EY035681 (ARD), UH3 EY031809 (ARD): Core Grant for Vision Research EY01792 all from NEI/NIH; Unrestricted Grant to the Department and Physician-Scientist Award both from Research to Prevent Blindness.

Conflicts of interest

The authors declare no conflicts of interest.