Safety and feasibility of mesenchymal stem cell therapy in patients with aqueous deficient dry eye disease

Full text

The Ocular Surface 19 (2021) 43–52
Available online 28 November 2020
1542-0124/© 2020 Published by Elsevier Inc.
Safety and feasibility of mesenchymal stem cell therapy in patients with
aqueous decient dry eye disease
Michael Møller-Hansen
a
,
*
, Ann-Cathrine Larsen
a
, Peter Bjerre Toft
a
,
Charlotte Duch Lynggaard
b
, Camilla Schwartz
c
, Helle Bruunsgaard
d
,
Mandana Haack-Sørensen
e
, Annette Ekblond
e
, Jens Kastrup
e
, Steffen Heegaard
a
a
Department of Ophthalmology, Rigshospitalet-Glostrup, University of Copenhagen, Denmark
b
Department of Otorhinolaryngology, Head and Neck Surgery and Audiology, Rigshospitalet, University of Copenhagen, Denmark
c
Department of Diagnostic Radiology, Rigshospitalet-Glostrup, University of Copenhagen, Denmark
d
Department of Clinical Immunology, Rigshospitalet, And Department of Clinical Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark
e
Cardiology Stem Cell Centre, The Heart Centre, Rigshospitalet, University of Copenhagen, Denmark
ARTICLE INFO
Keywords:
Dry eye disease
Stem cell therapy
Mesenchymal stem cells
Regenerative medicine
ABSTRACT
Purpose: To evaluate the safety and feasibility of injecting allogeneic adipose-derived mesenchymal stem cells
(ASCs) into the lacrimal gland (LG) as a treatment of aqueous decient dry eye disease (ADDE).
Methods: In this open-label, 5-visit clinical trial (baseline, treatment and weeks 1, 4 and 16) seven subjects with
ADDE received one transconjunctival injection of allogeneic ASCs into the LG in one eye. The ASC product
contained 22 million ASCs/ml and the injected volume was maximally 50% of the LG volume as determined on
magnetic resonance imaging (MRI). Treatment related adverse events (AEs) were assessed at each visit (primary
endpoint). Ocular Surface Disease Index (OSDI), tear osmolarity, tear lm breakup time (TBUT), corneal staining
(Oxford grade) and Schirmer’s I test were assessed at each timepoint.
Results: No AEs related to the study treatment were observed. Mean follow-up time was 126 days after treatment.
The mean OSDI score decreased from 58.9 ±20.6 at baseline to 34.1 ±21.6 (p <0.002). In the study eye mean
tear osmolarity decreased from 312.9 ±10.4 to 291.6 ±10.9 mosm/l (p <0.002), mean TBUT increased from
3.7 ±1.5 to 7.1 ±1.9 s (p <0.002), mean Schirmer’s I test increased from 4.6 ±0.7 to 8.1 ±3.1 mm/5 min (p <
0.03), while mean Oxford grade showed a trend towards a decrease from 2.4 ±0.7 to 1.3 ±1 (p <0.10).
Conclusion: Our trial suggests that injection of allogeneic ASCs into the LG is a safe and feasible treatment of
severe ADDE. A randomized placebo-controlled trial aimed at elucidating the therapeutic effect of allogeneic
ASCs in a larger patient cohort from our research group is currently underway.
1. Introduction
Dry eye disease (DED) is a common problem seen in patients all over
the world with a reported prevalence of 5–50% [1]. The prevalence of
DED in a Danish population has been reported to be 11% and increasing
with age [2]. Symptoms of DED include ocular discomfort and blurred
vision, which negatively impact visual function and quality of life. DED
is subdivided into evaporative dry eye (EDE) with excessive evaporation
from the tear lm and aqueous-decient dry eye (ADDE) with reduced
lacrimal secretion from the lacrimal gland (LG), though often patients
have a combination of both [3]. In the western world the most common
cause of ADDE is Sj¨
ogren’s syndrome (SS) which is a chronic
autoimmune disorder characterized by inammation and lymphocytic
inltration in the affected exocrine glands [4]. This exocrinopathy can
be encountered alone in primary SS (pSS) or in the presence of another
autoimmune disorder such as rheumatoid arthritis (secondary SS).
Presently, a diagnosis of SS relies on a combination of clinical, histo-
logical, and immunological ndings, in combination with relevant
subjective symptoms (dry eyes and/or mouth) as stated in the “2016
American College of Rheumatology/European League Against Rheu-
matism classication criteria for primary Sj¨
ogren’s syndrome” [5].
Current treatment options for ADDE are symptomatic as a curative
treatment does not exist [6].
Mesenchymal stromal - or stem cells (MSCs) are multipotent stem
* Corresponding author.
E-mail address: michael.moeller-hansen@regionh.dk (M. Møller-Hansen).
Contents lists available at ScienceDirect
The Ocular Surface
journal homepage: www.elsevier.com/locate/jtos
https://doi.org/10.1016/j.jtos.2020.11.013
Received 1 September 2020; Received in revised form 6 November 2020; Accepted 24 November 2020

The Ocular Surface 19 (2021) 43–52
44
cells with the capacity to differentiate into various cell types and have
been shown to reduce inammation and enhance tissue repair after
transplantation in vivo [7,8]. The immunomodulatory effect of MSCs on
the immune system can be mediated both through soluble factors and
cell to cell interactions, however, the paracrine signaling pathways are
considered the key mechanisms by which MSCs inuence other cells.
Several studies have reported therapeutic benets of MSCs trans-
plantation even though long-term engraftment of these cells could not
be detected [9]. Allogeneic MSCs has the potential advantage as an
“off-the-shelf” therapeutic agent, avoiding the cost of individual tissue
collection and cell culture associated with autologous MSCs. Also the
function of autologous MSCs has been proposed to be impaired in pa-
tients with comorbidities or advanced age [10]. An extensive number of
clinical trials with allogeneic MSCs from various sources and targeting
different diseases have been conducted and no adverse events related to
an anti-donor immune response have been reported [11]. Adipose
tissue-derived MSCs (ASCs) have gained considerable attention, as ASCs
are readily available from the abdominal fat where it is easily collected
and expanded ex vivo [12].
The substantial immunoregulatory and tissue reparative functions of
MSCs in the setting of ocular surface disease have been clearly demon-
strated [13]. Although the immunoregulatory mechanisms of MSCs in
DED have not been clearly delineated, preliminary reports suggest that
MSCs may protect the ocular surface against autoimmune mediated
inammation [14,15]. As in humans, canines are at risk of developing an
immune-mediated inammatory disease targeting the LGs which make
canines a superior animal model of ADDE [16]. Two trials evaluating
injection of allogeneic ASCs around the canine LG have been published
and both found that this treatment was safe, increased tear production,
and improved ocular surface signs up to 12 months after a single
treatment [17,18].
2. Materials and methods
2.1. Patient enrolment
We performed an open-label clinical trial to assess the safety and
feasibility of injecting allogeneic ASCs into the LG in patients with
ADDE. The trial was conducted according to the principles of the
Declaration of Helsinki, the ICH-GCP Guideline, and was monitored by
the GCP unit in the Capital Region of Denmark. The trial was approved
by the Danish National Committee on Health Research Ethics, the
Danish Medicines Agency (EudraCT no. 2018-003387-31), and was
registered as a clinical trial at ClinicalTrials.gov (NCT03878628).
The inclusion criteria were: (1) Age >18 years (2) Ocular Surface
Disease Index (OSDI) score >30 (3) Schirmer’s I test 2–5 mm/5 min (4)
tear lm break-up time (TBUT) <10 s. If both eyes fullled the inclusion
criteria, the study eye was determined as the eye with the lowest
Schirmer’s I test.
The exclusion criteria were: (1) Previously established allergies to
oxybuprocain or dimethyl sulfoxide (DMSO) (2) Previous treatment
with ASCs or other stem cell products in the LG(s) (3) Reduced immune
response (e.g. HIV positive) (4) Pregnancy or planned pregnancy within
the next two years (5) Breastfeeding (6) Treatment with an anticoagu-
lant that cannot be stopped during the intervention period (7) Treatment
with systemic medication known to reduce tear production: anxiolytics,
antipsychotics, and inhaled steroids (8) Topical treatment with eye
drops other than lubricants (9) Any other disease/condition judged by
the investigator to be grounds for exclusion, such as infection in or
around the eye.
2.2. Magnetic resonance imaging protocol
For safety reasons we decided to measure the volume of the lacrimal
Fig. 1. (A): Magnetic resonance image of the left orbit in the anterior-posterior projection at baseline in subject no. 3. The dotted line outlines the left lacrimal gland
(B): Transconjunctival injection of adipose-derived mesenchymal stem cells in subject no. 3.
Table 1
Subject demographic and disease characteristics at baseline.
Subject Sex Age,
years
Etiology LG volume (cm3),
study eye
Injection dose
(ml)
1 Female 55 pSS 0,2 0,1
2 Female 75 sSS 0,2 0,1
3 Female 43 sSS 0,4 0,2
4 Female 64 sSS 0,26 0,1
5 Female 56 sSS 0,5 0,2
6 Female 60 pSS 0,56 0,2
7 Female 66 sSS 0,33 0,1
pSS, primary Sj¨
ogren’s Syndrome; sSS, secondary Sj¨
ogren’s Syndrome; LG,
lacrimal gland.
M. Møller-Hansen et al.

The Ocular Surface 19 (2021) 43–52
45
glands in each study participant with magnetic resonance imaging (MRI)
at baseline and to inject a volume of the ASC suspension corresponding
to maximally 50% of the LG volume. The MRI protocol consisted of a 3D
T1 TFE sequence of the orbit in each subject. The volume of the LGs was
estimated by measuring the dimensions of the LG in the coronal and
anterior-posterior projection (Fig. 1A).
2.3. Clinical and laboratory assessment
The subjects were examined at baseline (maximum 30 days before
treatment), 30 min after treatment and at weeks 1, 4 and 16 after
treatment. Due to the COVID-19 pandemic ve visits (two week 4 and
three week 16) were conducted by telephone to register AEs and ll out
the OSDI questionnaire. The eye examination as per protocol for each
visit was postponed and performed as soon as possible according to the
guidelines regarding elective out-patient activities given by the Danish
Health Authority. Of these ve visits only the week 4 visit in subject no.
6 was not conducted due to COVID-19 measures. All subjects completed
the full examination at the nal follow-up visit after a minimum of 16
weeks. At each follow-up, adverse events were registered and symptoms
of DED were assessed with the validated 12-item Ocular Surface Disease
Index (OSDI) questionnaire (range 0–100: 0–12 is normal, 13–22 mild,
23–32 moderate, and ≥33 severe DED) [19]. The ophthalmological
examination was performed by the same experienced investigator at
each follow-up in the following order as recommend in the TFOS DEWS
II report [20]: Tear osmolarity measurement was performed using the
TearLab Osmolarity System (TearLab Corp., San Diego, CA) where a
positive result is normally considered to be ≥308 mOsm/l or an inter-
ocular difference >8 mOsm/l. 1 drop of sterile saline was then applied to
a sterile 1 mg uorescein sodium strip (I-DEW FLO, Entod Research Cell
UK Ltd., London, UK) which was applied to the inferior conjunctival
fornix of each eye. The tear break-up time (TBUT) was recorded in the
blue light on the slit lamp as the time between blinking and the rst dry
spot. Ocular surface staining was graded according to the Oxford scheme
(score 0–5) within 3 min after instillation of uorescein. Schirmer’s I test
without anesthesia was performed using sterile, standardized Schirmer
strips (I-DEW Tearstrips, Entod Research Cell UK Ltd., London, UK). The
strip was bent at the 0 mm mark and placed carefully over the lower lid
margin 1/3 towards the temporal angle of the lids. The strip remained in
place for 5 min where the wetting of the strips was measured using the
printed millimeter scale.
A potential immune response to the transplantation of allogeneic
ASCs in all subjects was evaluated by detection of IgG human leukocyte
antigen (HLA) antibodies in patient serum before treatment and at each
follow-up. Luminex based screening assay for HLA antibodies was per-
formed (LabScreen Mixed, One Lambda, Inc., Thermo Fisher, Canoga
Park, CA.) where a positive screening was dened by a normalized
background ratio. In the case of a positive screening test result, a
Table 2
Safety prole with description of the adverse events and severe adverse events during the study period.
Adverse events (AEs) Baseline Treatment* 1 week 4 weeks 16 weeks None
Study treatment related AE None
Injection related AE
Pain at injection site, n (%) 3 (43%) 3 (43%) 1 (14%)
Grade 1, subject 2
Grade 2, subject 2, 5 2, 5, 6
Grade 3, subject. 6
Periorbital edema, n (%) 1 (14%) 3 (43%) 5 (71%) 1 (14%) 2 (29%)
Grade 1, subject. 1 3, 6, 7 1, 2
a
, 3, 4, 6 2
a
1, 2
a
Grade 2, subject
Grade 3, subject.
Ocular discomfort, n (%) 6 (86%) 7 (100%) 5 (71%) 2 (29%) 2 (29%)
Grade 1, subject. 1, 2, 3, 4 3, 4, 6, 7 3, 5, 7 2 3
Grade 2, subject. 5, 6 1, 2, 5 1 6
Grade 3, subject 6
Periorbital paresthesia, n (%) 1 (14%)
Grade 1, subject 1
Grade 2, subject
Grade 3, subject
Blurred vision, n (%) 3 (43%) 1 (14%)
Grade 1, subject 4 3
Grade 2, subject 2, 5
Grade 3, subject
Infection at injection site None
Bleeding from injection site None
Eyelid function disorder None
Other
Flu-like symptoms, n (%) 1 (14%) 2 (29%)
Grade 1, subject 1
b
5
c
Grade 2, subject 7
d
Grade 3, subject.
Life threatening None
Non-life threatening
Related to study treatment None
Not related to study treatment
Hospitalization, subject 1
e
Hospitalization, subject. 1
f
*within 30 min after treatment.
a
; head trauma against the study eye before 4-weeks follow-up.
b
; pneumonia.
c
; COVID-19 symptoms.
d
; sore throat, COVID negative.
e
; cholecystitis.
f
; general malaise.
M. Møller-Hansen et al.

The Ocular Surface 19 (2021) 43–52
46
Luminex based single antigen bead assay for detection of specic HLA
antibodies was performed (LabScreen SA, One Lambda, Inc., Thermo
Fisher, Canoga Park, CA). Cut-off for positivity in the single antigen
assay was dened as mean uorescence intensity (MFI) >1000. A vir-
tual crossmatching was performed to evaluate if any donor specic
antibodies (DSA) were present before or developed de novo after
treatment. HLA typing and antibody detection and identication ana-
lyses were accredited by European Federation for Immunogenetics.
2.4. Preparation of allogeneic ASC product
The ASC product was manufactured by Cardiology Stem Cell Centre
(CSCC), University Hospital Copenhagen, Rigshospitalet, as previously
described [11,21–23]. CSCC holds a manufacturing authorization for
advanced therapy investigational medicinal products and a tissue
establishment authorization as issued and inspected by Danish Medi-
cines Agency and Danish Patient Safety Authority, respectively.
In short, lipoaspirates were obtained from three healthy female do-
nors (25–34 years old). Donor eligibility was determined by donor in-
terviews and testing for infectious disease markers: human
immunodeciency virus (HIV), hepatitis B and C, syphilis and human.
T-cell lymphotropic virus (HTLV I/II) within 30 days prior to lipo-
suction. In addition, blood samples were drawn on the day of donation
for repeated serology and nucleic acid testing (NAT) of HIV, and hepa-
titis B and C. ASCs were expanded in automated closed bioreactor
systems (Quantum Cell Expansion System, Terumo BCT) with human
platelet lysate as a growth supplement (Sexton Biotechnologies).
Expansion of ASCs was performed in a two-passage process. Primary
expansion of the stromal vascular fraction produced rst passage ASCs,
which were further expanded in a second passage to produce ASCs for
the nal product. The nal product was cryopreserved in CellSeal vials
(Sexton Biotechnologies) as a customized formulation of the product
CSCC_ASC holding 22 million ASCs per ml with a total volume of 1.2 ml
per vial. The excipient was CryoStor CS10 (BiolifeSolutions). The
product was stored below −180
∘
C in nitrogen dry-storage until clinical
use. Release criteria were viral safety (donor serology/NAT), sterility
(including bacteria, fungus, mycoplasmas and endotoxins), cell number
and viability (>80%) and immunophenotypical characterization of cells
by ow cytometry (stable positive markers CD90, CD105, CD73 >80%
and negative markers <3% CD45, <5% human leukocyte antigen
(HLA)-DR). 24-month stability of release criteria and cell function dur-
ing storage was documented. All donors were HLA-A, -B, –C, -DRB1,
-DRB3/4/5, -DQA1, -DQB1, -DPA1, and -DPB1 typed by qPCR (Linkseq,
Linkage Biosciences, One Lambda, Inc., Thermo Fisher, Canoga Park,
CA). The product was produced from 3 donors, but every batch/treat-
ment unit was based on one donor only.
2.5. Cell therapy protocol
Each subject received one transconjunctival injection of the alloge-
neic ASC product into the LG in one eye. The injection volume varied
corresponding to a maximum of 50% of the LG volume as assessed on
MRI. All procedures were performed by the same experienced consultant
eye surgeon using sterile technique in an outpatient setting. The ocular
surface was prepared for injection using two drops of oxybuprocain
0.4% with 30 s intervals, then 2 drops of Povidone-iodine 5% with 30 s
intervals, and nally, the upper eyelid was retracted by the surgeon, the
palpebral part of the LG was identied and the conjunctival injection
site was anaesthetized in sterile conditions using a cotton swap soaked in
a cocaine 10% solution for a minimum of 2 min. The ASC product comes
frozen as a ready-to-use cell product in which the ASCs are suspended in
the excipient medium CryoStor CS10. Within 1 h from thawing bed side
in a 37
⸰
C water bath, the designated volume of the ASC product was
extracted into a 1 ml syringe. The ASC product was then injected
transconjunctivally into the lacrimal gland as previously described by
Montoya et al. [24] using a 10 mm long 30G needle (Fig. 1B). The in-
jection was administered slowly over the course of approximately 10 s.
At the end of the injection, the needle was kept in the lacrimal gland for
10 s with slight pressure on the syringe piston to avoid backow of the
injected solution.
2.6. Outcome measures
The primary outcome measure was safety evaluated as the study
treatment related adverse events (AEs) according to the Common Ter-
minology Criteria for Adverse Events (CTCAE) [25]. The AEs were
classied as either related to the study treatment or not. Unrelated AEs
were then classied as either “procedure related” or “other causes”.
Secondary outcome measures were changes in OSDI score, tear osmo-
larity, TBUT, Oxford grade, Schirmer’s I test, and development of donor
specic HLA antibodies.
2.7. Statistical analysis
Adverse events (AEs) were summarized using descriptive statistics.
The results addressing our secondary outcomes were calculated as
means ±standard deviation (SD) at each time point and analyzed using
paired t-tests comparing the values at baseline and after 16 weeks
follow-up. The same was performed for differences between study eye
and fellow eye at each time point. Differences were considered statisti-
cally signicant if the two-sided p-value was less than 0.05.
Table 3
Secondary outcome measures at baseline and after 16 weeks follow-up.
Outcome measure Value P
OSDI score 0.002†
Baseline 58.9 ±
20.6
Last follow-up 34.1 ±
21.6
Study
eye
P Fellow
eye
p p
(interocular)
Visual acuity
(logMar), mean ±
SD
0.10 0.17
Baseline 0.1 ±
0.1
0.09 ±
0.11
0.6
Follow-up 0.04 ±
0.13
0.06 ±
0.12
0.36
Tear osmolarity
(mosm/l), mean ±
SD
0.002y0.34
Baseline 312.9 ±
10.4
306.3 ±
10.9
0.2
Follow-up 291.6 ±
10.9
305.7 ±
11.2
0.04§
TBUT (s), mean ±SD 0.002y0.34
Baseline 3.7 ±
1.5
3.7 ±
0.9
1.0
Follow-up 7.1 ±
1.9
4.7 ±
1.6
0.02§
Corneal staining
(Oxford grade),
mean ±SD
0.10 0.36
Baseline 2.4 ±
0.7
2.3 ±
0.9
0.36
Follow-up 1.3 ±1 1.9 ±
0.8
0.1
Schirmer’s I test (mm/
5 min), mean ±SD
0.03y0.17
Baseline 4.6 ±
0.7
4.6 ±
3.2
1.0
Follow-up 8.1 ±
3.1
6.9 ±
4.5
0.25
LG, lacrimal gland; OSDI, Ocular Surface Disease Index; TBUT, tear break-up
time. †p <0,05 from baseline to last follow-up. §p <0,05 between study eye
and fellow eye.
M. Møller-Hansen et al.

The Ocular Surface 19 (2021) 43–52
47
3. Results
Seven subjects signed the informed consent form and were screened
for this trial. All seven subjects fullled the inclusion criteria and no
exclusion criteria and had prior been diagnosed with either primary (n
=2) or secondary Sj¨
ogren’s syndrome (n =5). Both LGs were identied
on the MRI in all subjects. The mean volume of the LGs in the study eyes
was 0.35 ±0.13 ml. The ASC dose was either 0.1 ml (n =4) or 0.2 ml (n
=3) corresponding to a dose of 2.2 ×10
6
or 4.4 ×10
6
ASCs per LG or a
range from 8.46 ×10
6
to 11 ×10
6
ASCs per LG volume in ml. The
demographic and disease characteristics at baseline are listed in Table 1.
Within a mean follow-up of 126 days after treatment, no adverse
events (AEs) related to the study treatment were observed. The pro-
cedure related AEs included pain at injection site, periorbital edema,
increase in ocular discomfort, blurred vision and periorbital paresthesia.
Two serious adverse events (SAEs) were recorded during follow-up, both
in subject no. 1. None of the SAEs were life threatening or related to the
study treatment. AEs and SAEs are listed in Table 2.
Several secondary outcome measures changed signicantly during
the follow-up period (Table 3). Most notable is the 45% decrease in
mean OSDI score from baseline to last follow-up (58.9 ±20.6 to 34.1 ±
21.6, p <0.002) (Fig. 2). No signicant differences presented between
mean measurements in the study and fellow eye at baseline, while mean
tear osmolarity (p <0.04) and TBUT (p <0.02) showed a signicant
difference in favor of the study eye at last follow-up.
In the study eye the mean tear osmolarity decreased (312.9 ±10.4 to
291.6 ±10.9 mosm/l, p <0.002), mean TBUT increased (3.7 ±1.5 to
7.1 ±1.9 s, p <0.002), and mean Schirmer’s I test increased (4.6 ±0.7
to 8.1 ±3.1 mm/5 min, p <0.03) signicantly. Mean Oxford grade
showed a trend toward a decrease (2.4 ±0.7 to 1.3 ±1, p <0.10)
(Fig. 3A–D). In the fellow eye none of the outcome measures changed
signicantly. The mean tear osmolarity remained unchanged (306.3 ±
10.9 to 305.7 ±11.2 mosm/l, p <0.34), while mean TBUT (3.7 ±0.9 to
4.71 ±1.6 s, p <0.18) and Schirmer’s I test increased insignicantly
(4.6 ±3.2 to 6.9 ±4.5 mm/5 min, p <0.17). Mean Oxford grade
decreased insignicantly (2.3 ±0.9 to 1.9 ±0.8, p <0,36) (Fig. 4A–D).
The results from Luminex screening and specic single antigen assay
for HLA antibodies in each subject are listed in Supplemental Table 4.
Two subjects were sensitized with pre-existing HLA antibodies including
DSA against HLA class II antigens before the study treatment. These HLA
antibodies persisted but were not boosted during the study period: at
baseline subject no. 3 had DSA against HLA-DR13 and HLA-DR52 with
MFI 1000–3000 and subject no. 7 had DSA against HLA-DQ8 with MFI
1000–3000. The latter subject developed de novo DSA against HLA-Cw9
with low MFI (<1500) after 4 weeks, which persisted after 16 weeks.
Additionally, subject no. 2 developed de novo DSA against HLA-B44
with low MFI (<1500) after 16 weeks. No subjects developed clinical
symptoms or inammatory signs indicating immunization during the
follow-up period. The de novo development of DSA had no inuence on
the secondary outcome measures in subject 2 and 7 compared to the
mean values of all subjects (Figs. 2–4).
4. Discussion
This is the rst clinical trial to evaluate injection of allogeneic ASCs
into the LG as a treatment of DED in humans. The data from this trial
suggests that the ASC treatment is safe and well tolerated. Signs and
symptoms of DED are often poorly correlated, which makes simulta-
neous, signicant changes in both signs and symptoms in clinical trials
of DED rare [26]. The results in this trial, however, showed a clinically
relevant response across signs and symptoms of ADDE due to either
primary or secondary Sj¨
ogren’s syndrome, albeit in a small number of
patients. The magnitude and rapidity of improvement in DED symptoms
is also notable, as 80% of the mean reduction in OSDI score achieved 16
weeks after treatment was even present after one-week follow-up
(Fig. 2). One can then speculate whether a cumulative effect of
consecutive treatments might be present. It seems that the objective
signs of DED respond to the treatment within weeks rather than days
(Fig. 3A–D).
Injecting too large a volume of the ASC suspension into the LG might
lead to an increase in pressure within the LG capsule with the risk of
damaging the glandular tissue. For safety reasons we decided to maxi-
mally inject a dose corresponding to 50% of the LG volume as estimated
with MRI at baseline. No adverse events assessed to be related to the
volume of the injected ASCs suspension was found.
No pre-treatment tissue type matchings between the donors and the
subjects were carried out. Two subjects had donor-specic HLA class II
antibodies at baseline and two subjects developed de novo donor-
Fig. 2. Changes in Ocular Surface Disease Index (OSDI) scores from baseline (day 0) to last follow-up (range: 112–147 days) in subject no. 1–7.
M. Møller-Hansen et al.

The Ocular Surface 19 (2021) 43–52
48
specic HLA class I antibodies with low MFI after treatment. However,
this seemed to have no inuence on the outcome measures which has
also been concluded in other MSCs studies [27].
DED usually presents as a bilateral disease although there can be
differences in disease severity between the eyes. Our results show a
signicant difference in tear osmolarity and TBUT in favor of the study
eye compared to the untreated fellow eye, however, a trend towards
improvement of the secondary outcome measures in the untreated eye is
also seen (Fig. 4A–D). It is well established, that the eyes do not function
as independent units but rather “communicate” with effects on the
function of the other eye, though the mechanism of this communication
between the eyes has not been clearly established [44]. For this reason,
the fellow eye cannot function as a perfect control to the study eye.
The injection of the ASC product stabilized the tear lm in the study
eye as seen in the signicant decrease in tear osmolarity and increase in
TBUT, indicating decreased inammatory activity on the ocular surface.
It has previously been shown that MSCs decrease pro-inammatory T-
cell responses and increase anti-inammatory T-cell responses [28].
Regulatory T cells (Tregs) secrete anti-inammatory cytokines such as
TGF-β and IL-10, which may play a role in mediating the response from
MSCs and accelerate LG repair [29]. The mechanism of how ASCs might
modulate the inammatory response in the LG following transplantation
of the ASC product in the present study, however, is speculative. One
feasible method to elucidate the anti-inammatory mechanism of ASCs
in future trials is the measurement of cytokine levels in tear uid [30].
The mechanism of action is particularly important when considering
Fig. 3. Study eye. Changes in secondary outcome measures from baseline (day 0) to last follow-up (range: 112–147 days) in subject no. 1–7. (A): Tear osmolarity (B):
Tear break-up time (C): Corneal staining according to the Oxford scheme (D): Schirmer’s I test.
M. Møller-Hansen et al.

The Ocular Surface 19 (2021) 43–52
49
which treatment population should be selected for a clinical trial. ASCs
appear to have potent immune regulatory actions that make ASCs
attractive for use in inammatory diseases. This is indeed the case in
DED, but also in corneal injury, allergic conjunctivitis, ocular graft-
versus-host disease, and following corneal transplantation. We hypoth-
esize that this treatment is more likely to have an effect in patients with
early phases of the various inammatory diseases than in later phases
with presence of e.g. cicatricial conjunctivitis. The morphological fea-
tures of the palpebral lobe of the main lacrimal gland might even be a
useful additional outcome measure for use in future trials [31].
Limitations of this trial include a small number of subjects and the
lack of placebo comparison and blinding. However, a double-blinded,
randomized clinical trial from our research group is already underway.
5. Conclusion
Our study found injection of allogeneic adipose-derived mesen-
chymal stem cells into the lacrimal gland safe, feasible and well toler-
ated. A single injection of allogeneic ASCs generally improved both signs
and symptoms in patients with severe ADDE due to either primary or
secondary Sj¨
ogren’s syndrome. The improvement of the secretory
function of the lacrimal gland was measured with a signicant increase
in tear break-up time and Schirmer’s I test and a decrease in tear os-
molarity in the study eye. Further studies aiming at elucidating the
therapeutic effect of allogeneic adipose-derived mesenchymal stem cells
and the downstream targets involved in the improved lacrimal gland
function in a larger patient cohort are warranted.
Fig. 3. (continued).
M. Møller-Hansen et al.

The Ocular Surface 19 (2021) 43–52
50
Funding/disclosures
This study was funded by grants from the Synoptik Foundation and
Fight for Sight, Denmark. In collaboration with Rigshospitalet and the
University of Copenhagen authors Michael Møller-Hansen, Ann-
Cathrine Larsen and Steffen Heegaard has led a patent application
“Stem cell therapy for lacrimal gland dysfunction” to the European
Patent Ofce (ref. 21943EP00). Mandana Haack-Sørensen, Annette
Ekblond and Jens Kastrup are inventors of the patent “Stem cell therapy
based on adipose-derived stem cells " (Publication WO 2017–068,140).
All authors met the ICMJE authorship criteria. No honoraria or pay-
ments were made for authorship.
Fig. 4. Fellow eye. Changes in secondary outcome measures from baseline (day 0) to last follow-up (range: 112–147 days) in subject no. 1–7. (A): Tear osmolarity
(B): Tear break-up time (C): Corneal staining according to the Oxford scheme (D): Schirmer’s I test.
M. Møller-Hansen et al.

The Ocular Surface 19 (2021) 43–52
51
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.jtos.2020.11.013.
References
[1] Stapleton F, Alves M, Bunya VY, Jalbert I, Lekhanont K, Malet F, et al. TFOS DEWS
II epidemiology report. Ocul Surf 2017;15:334–65. https://doi.org/10.1016/j.
jtos.2017.05.003.
[2] Bjerrum KB. Keratoconjunctivitis sicca and primary Sj¨
ogren’s syndrome in a Danish
population aged 30-60 years. Acta Ophthalmol Scand 1997;75:281–6.
[3] Bron AJ, de Paiva CS, Chauhan SK, Bonini S, Gabison EE, Jain S, et al. TFOS DEWS
II pathophysiology report. Ocul Surf 2017;15:438–510. https://doi.org/10.1016/j.
jtos.2017.05.011.
[4] Gomes JAP, Azar DT, Baudouin C, Efron N, Hirayama M, Horwath-Winter J, et al.
TFOS DEWS II iatrogenic report. Ocul Surf 2017;15:511–38. https://doi.org/
10.1016/j.jtos.2017.05.004.
[5] Shiboski CH, Shiboski SC, Seror R, Criswell LA, Labetoulle M, Lietman TM, et al.
American College of Rheumatology/European League against Rheumatism
classication criteria for primary Sj¨
ogren’s syndrome: a consensus and data-driven
methodology involving three international patient cohorts. Ann Rheum Dis 2016;
76:9–16. https://doi.org/10.1136/annrheumdis-2016-210571. 2017.
Fig. 4. (continued).
M. Møller-Hansen et al.

The Ocular Surface 19 (2021) 43–52
52
[6] Jones L, Downie LE, Korb D, Benitez-Del-Castillo JM, Dana R, Deng SX, et al. TFOS
DEWS II management and therapy report. Ocul Surf 2017;15:575–628. https://doi.
org/10.1016/j.jtos.2017.05.006.
[7] Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, et al.
Multilineage potential of adult human mesenchymal stem cells. Science 1999;284:
143–7.
[8] Le Blanc K, Davies LC. MSCs—cells with many sides. Cytotherapy 2018;20:273–8.
https://doi.org/10.1016/j.jcyt.2018.01.009.
[9] Caplan A. Why are MSCs therapeutic? New data: new insight. J Pathol 2009;217:
318–24. https://doi.org/10.1002/path.2469.
[10] Stolzing A, Jones E, McGonagle D, Scutt A. Age-related changes in human bone
marrow-derived mesenchymal stem cells: consequences for cell therapies. Mech
Ageing Dev 2008;129:163–73. https://doi.org/10.1016/j.mad.2007.12.002.
[11] Kastrup J, Haack-Sørensen M, Juhl M, Harary Søndergaard R, Follin B, Drozd
Lund L, et al. Cryopreserved off-the-shelf allogeneic adipose-derived stromal cells
for therapy in patients with ischemic heart disease and heart failure-A safety study.
Stem Cells Transl Med 2017;6:1963–71. https://doi.org/10.1002/sctm.17-0040.
[12] Rodriguez A-M, Elabd C, Amri E-Z, Ailhaud G, Dani C. The human adipose tissue is
a source of multipotent stem cells. Biochimie 2005;87:125–8. https://doi.org/
10.1016/j.biochi.2004.11.007.
[13] Sahu A, Foulsham W, Amouzegar A, Mittal SK, Chauhan SK. The therapeutic
application of mesenchymal stem cells at the ocular surface. Ocul Surf 2019;17:
198–207. https://doi.org/10.1016/J.JTOS.2019.01.006.
[14] Lee MJ, Ko AY, Ko JH, Lee HJ, Kim MK, Wee WR, et al. Mesenchymal stem/stromal
cells protect the ocular surface by suppressing inammation in an experimental dry
eye. Mol Ther 2015;23:139–46. https://doi.org/10.1038/mt.2014.159.
[15] Aluri HS, Samizadeh M, Edman MC, Hawley DR, Armaos HL, Janga SR, et al.
Delivery of bone marrow-derived mesenchymal stem cells improves tear
production in a mouse model of Sj¨
ogren’s syndrome. Stem Cell Int 2017;2017:
1–10. https://doi.org/10.1155/2017/3134543.
[16] Quimby FW, Schwartz RS, Poskitt T, Lewis RM. A disorder of dogs resembling
Sj¨
ogren’s syndrome. Clin Immunol Immunopathol 1979;12:471–6.
[17] Villatoro AJ, Fern´
andez V, Claros S, Rico-Llanos GA, Becerra J, Andrades JA. Use of
adipose-derived mesenchymal stem cells in keratoconjunctivitis sicca in a canine
model. BioMed Res Int 2015;2015:1–10. https://doi.org/10.1155/2015/527926.
[18] Bittencourt MKW, Barros MA, Martins JFP, Vasconcellos JPC, Morais BP,
Pompeia C, et al. Allogeneic mesenchymal stem cell transplantation in dogs with
keratoconjunctivitis sicca. Cell Med 2016;8:63–77. https://doi.org/10.3727/
215517916X693366.
[19] Schiffman RM, Christianson MD, Jacobsen G, Hirsch JD, Reis BL. Reliability and
validity of the ocular surface disease Index. Arch Ophthalmol 2000;118:615.
https://doi.org/10.1001/archopht.118.5.615.
[20] Wolffsohn JS, Arita R, Chalmers R, Djalilian A, Dogru M, Dumbleton K, et al. TFOS
DEWS II diagnostic methodology report. Ocul Surf 2017;15:539–74. https://doi.
org/10.1016/j.jtos.2017.05.001.
[21] Haack-Sørensen M, Juhl M, Follin B, Harary Søndergaard R, Kirchhoff M,
Kastrup J, et al. Development of large-scale manufacturing of adipose-derived
stromal cells for clinical applications using bioreactors and human platelet lysate.
Scand J Clin Lab Invest 2018;78:293–300. https://doi.org/10.1080/
00365513.2018.1462082.
[22] Paitazoglou C, Bergmann MW, Vrtovec B, Chamuleau SAJ, van Klarenbosch B,
Wojakowski W, et al. Rationale and design of the European multicentre study on
stem cell therapy in IschEmic non-treatable cardiac diseasE (SCIENCE). Eur J Heart
Fail 2019;21:1032–41. https://doi.org/10.1002/ejhf.1412.
[23] Kastrup J, Schou M, Gustafsson I, Nielsen OW, Møgelvang R, Kofoed KF, et al.
Rationale and design of the rst double-blind, placebo-controlled trial with
allogeneic adipose tissue-derived stromal cell therapy in patients with ischemic
heart failure: a phase II Danish multicentre study. Stem Cell Int 2017;2017:1–8.
https://doi.org/10.1155/2017/8506370.
[24] Montoya FJ, Riddell CE, Caesar R, Hague S. Treatment of gustatory
hyperlacrimation (crocodile tears) with injection of botulinum toxin into the
lacrimal gland. Eye 2002;16:705–9. https://doi.org/10.1038/sj.eye.6700230.
[25] Cancer Institute N. Common Terminology criteria for adverse events (CTCAE) v5.0.
2017.
[26] Novack GD, Asbell P, Barabino S, Bergamini MVW, Ciolino JB, Foulks GN, et al.
TFOS DEWS II clinical trial design report. Ocul Surf 2017;15:629–49. https://doi.
org/10.1016/j.jtos.2017.05.009.
[27] Kot M, Baj-Krzyworzeka M, Szatanek R, Musiał-Wysocka A, Suda-Szczurek M,
Majka M. The importance of HLA assessment in “off-the-shelf” allogeneic
mesenchymal stem cells based-therapies. Mol Sci 2019. https://doi.org/10.3390/
ijms20225680.
[28] Khalili S, Liu Y, Kornete M, Roescher N, Kodama S, Peterson A, et al. Mesenchymal
stromal cells improve salivary function and reduce lymphocytic inltrates in mice
with Sj¨
ogren’s-like disease. PloS One 2012;7. https://doi.org/10.1371/journal.
pone.0038615.
[29] Singer NG, Caplan AI. Mesenchymal stem cells: mechanisms of inammation. Annu
Rev Pathol 2011;6:457–78. https://doi.org/10.1146/annurev-pathol-011110-
130230.
[30] Shetty R, Sethu S, Chevour P, Deshpande K, Pahuja N, Nagaraja H, et al. Lower
vitamin D level and distinct tear cytokine prole were observed in patients with
mild dry eye signs but exaggerated symptoms. Transl Vis Sci Technol 2016;5.
https://doi.org/10.1167/tvst.5.6.16.
[31] Singh S, Shanbhag SS, Basu S. Palpebral lobe of the human lacrimal gland:
morphometric analysis in normal versus dry eyes. Br J Ophthalmol 2020. https://
doi.org/10.1136/bjophthalmol-2020-316929. bjophthalmol-2020-316929.
M. Møller-Hansen et al.