Journal Pre-proof

Submandibular gland transplantation vs. minor salivary glands

transplantation for treatment of dry eye: A retrospective cohort study

Jia-Zeng Su , Bang Zheng , Zhen Wang , Xiao-Jing Liu ,

Zhi-Gang Cai , Lei-Zhang , Xin-Peng , Jun Wu , Xin-Hua Liu ,
Lan Lv , Guang-Yan Yu

PII: S0002-9394(22)00216-1
DOI: https://doi.org/10.1016/j.ajo.2022.05.019
Reference: AJOPHT 12248

To appear in: American Journal of Ophthalmology

Received date: September 25, 2021

Revised date: May 18, 2022
Accepted date: May 19, 2022

Please cite this article as: Jia-Zeng Su , Bang Zheng , Zhen Wang , Xiao-Jing Liu , Zhi-Gang Cai ,
Lei-Zhang , Xin-Peng , Jun Wu , Xin-Hua Liu , Lan Lv , Guang-Yan Yu , Submandibular gland
transplantation vs. minor salivary glands transplantation for treatment of dry eye: A retrospective cohort
study, American Journal of Ophthalmology (2022), doi: https://doi.org/10.1016/j.ajo.2022.05.019

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.

© 2022 Published by Elsevier Inc.

 1 Submandibular gland transplantation vs. minor salivary glands

2 transplantation for treatment of dry eye: A retrospective cohort

3 study

4 Jia-Zeng Su1, Bang Zheng2, Zhen Wang3, Xiao-Jing Liu1, Zhi-Gang Cai1, Lei-

5 Zhang1, Xin-Peng1, Jun Wu4, Xin-Hua Liu5, Lan Lv6**, Guang-Yan Yu1**
1
6 Department of Oral and Maxillofacial Surgery, Peking University School and
7 Hospital of Stomatology, National Clinical Research Center for Oral Diseases,
8 National Engineering Laboratory for Digital and Material Technology of Stomatology,
9 Beijing 100081, P. R. China
2
10 School of Public Health, Imperial College London, London SW7 2AZ, UK
3
11 Department of Stomatology, Capital Medical University Affiliated Beijing Friendship
12 Hospital, Beijing, 100050, P. R. China
4
13 Department of Ophthalmology, Affiliated Beijing Bo Ai Hospital, Capital University of
14 Medical Science, Beijing, 100068, P. R. China.
5
15 The First People’s Hospital of Jinzhong, Shanxi Province, 030600, P. R. China
6
16 Department of Ophthalmology, Affiliated Beijing Tong Ren Hospital, Capital
17 University of Medical Science, Beijing 100730, P. R. China
18

19 ** Corresponding authors:
20 Guang-Yan Yu, Department of Oral and Maxillofacial Surgery, Peking University
21 School of Stomatology, 22 Zhong Guan Cun South St. Beijing 100081, P. R. China;
22 Tel: 86-10-82195245; Fax: 86-10-62173402; E-mail: gyyu@263.net
23 Lan Lv, Department of Ophthalmology, Affiliated Beijing Tong Ren Hospital, Capital
24 University of Medical Science, Beijing 100730, P. R. China; Tel: 86-10-82195992;
25 Fax: 86-10-62173402; E-mail: tryklvlan@126.com

26

27 Short title: Salivary glands transplantations for dry eye disease

28

1
 1 Highlights

2  Submandibular gland transfer is indicated for end-stage refractory dry eye.

3  Minor salivary glands transfer is indicated for less severe refractory dry eye.

4  Minor salivary glands transfer is not indicated for end-stage refractory dry eye.

6 Abstract

7 Purpose: To compare submandibular gland (SMG) transplantation with minor

8 salivary glands (MSGs) transplantation for the treatment of different dry eye diseases

9 (DED).

10 Design: Retrospective clinical cohort study.

11 Methods: Seventy-three refractory DED eyes were divided into three groups. Group

12 A: 35 end-stage DED eyes who underwent SMG transplantation. Group B: 20 end-

13 stage DED eyes with MSGs transplantation. Group C: 18 non-end-stage DED eyes

14 with MSGs transplantation. Schirmer test (ST), tear break-up time (TBUT), corneal

15 fluorescein staining (FL), and best-corrected visual acuity (BCVA) were measured

16 before and after surgery.

17 Results: The length of hospital stay, length of operation, and hospital fee were

18 significantly higher in group A than in group B or C. Eyes in group A showed the

19 most severe DED disease with preoperative ST, TBUT, FL, and BCVA of 0.36

20 mm/5min, 0.03 s, 10.97, and 0.11, respectively, which improved significantly to 20.23

2
 1 mm/5min, 1.74 s, 7.58, and 0.2 at > 2 years follow-up. Group B had similar baseline

2 data, and significant but limited improvement only in the ST (0.55 mm/5min to 3.79

3 mm/5min) and FL (11.10 to 9.58) after the operation. Group C had better baseline

4 ST, TBUT, FL, and BCVA of 0.89 mm/5min, 3.49 s, 1.83, and 0.81, respectively,

5 which improved significantly (except for BCVA) to 9.35 mm/5min, 9.08 s, 0.53, and

6 0.89 after MSGs transplantation.

7 Conclusion: SMG transplantation could be recommended to treat end-stage

8 refractory DED. MSGs transplantation may provide satisfying results for refractory

9 DED with relatively less severe impairment of the eye.

10
11 Keywords
12 Dry eye disease; salivary gland transplantation; submandibular gland transplantation; minor salivary
13 glands transplantation

14

15

16 INTRODUCTION

17 Due to its multifactorial etiology, the management of dry eye disease (DED) is

18 complicated. Recently, the Tear Film and Ocular Surface Society's Dry Eye

19 Workshop II (TFOS DEWS II) has proposed an evidence-based, multistaged

20 management algorithm to determine the most appropriate DED treatment for each

21 patient. The treatment is recommended based on subjective and objective severity

22 measurements. Health education, environmental modifications, ocular lubricants, tear

23 conservation and physical therapies, etc. are recommended firstly. When the above

3
 1 options are insufficient, prescription drugs, autologous/allogeneic serum eye drops

2 and therapeutic contact lenses are recommended. Surgeries are usually given for

3 refractory DED.1 Two surgical modalities of salivary gland transplantation,

4 microvascular submandibular gland (SMG) transplantation and minor salivary glands

5 (MSGs) transplantation, are listed as the last recommendations when other

6 treatments are inadequate.1 However, further instruction on choosing between these

7 two surgical approaches is still vague.1-20

8 SMG transplantation2-13 and MSGs transplantation14-20 provide spontaneous,

9 continuous, and endogenous lubrication with saliva as a tear substitute. Both

10 treatments can improve severe symptoms and/or signs of dry eye, including

11 Schirmer test score (ST), tear break-up time (TBUT), and some other ocular-surface

12 features. However, besides the above-listed similarities, these methods differ in

13 surgical complexity or technical requirement, surgical trauma, secretory patterns,

14 amounts of lubrication, and surgical complications. Besides, there is individual

15 heterogeneity among these refractory DED patients, and the etiology and severity of

16 the disease in these patients may vary. Taking into consideration significant

17 differences between the two surgeries and the heterogeneity among the refractory

18 DED patients, when salivary gland transplantation is indicated according to the

19 current algorithm, an evidence-based choice between SMG transplantation and

20 MSGs transplantation is needed.

21 In this study, we compared the treatment effect of SMG transplantation and

4
 1 MSGs transplantation in different refractory DED patients. We also analyzed surgical

2 complications and treatment costs. The current study provides evidence-based

3 indications for SMG transplantation and MSGs transplantation for the treatment of

4 severe refractory DED.

6 MATERIALS AND METHODS

7 Patients

8 Consecutive patients diagnosed with DED who underwent salivary gland

9 transplantation in Peking University School of Stomatology between June 2010 and

10 October 2018 were included in this retrospective clinical cohort study. The study was

11 approved by the Ethics Committee of the Peking University School of Stomatology

12 (PKUSSIRB - 202163043) and designed and carried out in full accordance with the

13 World Medical Association Declaration of Helsinki. All patients provided informed

14 consent to join the study.

15 Indications and contraindications

16 Indications for surgeries included persistently pronounced symptoms of dry eye and

17 failure of other previous ophthalmologic treatments, along with a ST value of <2 mm,

18 a TBUT value of <5 seconds, and positive fluorescence staining of the cornea during

19 ophthalmologic evaluation. Contraindications were Sjőgren’s syndrome or obvious

20 symptoms of xerostomia.8,13

5
 1 Grouping

2 Patients comprised three groups (Group A, B, and C) based on their surgical

3 modalities, the severity, and the etiologies of the DED (Table 1). DED patients

4 secondary to cicatrizing conjunctivitis and meeting the level-4 grade of dry eye

5 severity grading scheme in the 2007 International Dry Eye WorkShop (DEWS)

6 criteria21 who underwent SMG transplantation were included in group A; level-4 DED

7 patients secondary to cicatrizing conjunctivitis who underwent MSGs transplantation

8 were included in group B. The DED cases secondary to non-cicatrizing conjunctivitis

9 and not meeting the level-4 grade in the DEWS criteria were included in group C.

10 The surgical modality was mainly selected according to the patient’s intention after

11 surgical techniques have been explained in detail.

12 The causes of cicatrizing conjunctivitis included Stevens-Johnson syndrome

13 (SJS), mucous membrane pemphigoid (MMP), and graft versus host disease

14 (GVHD). Based on the DEWS dry eye severity grading scheme21, the inclusion

15 criteria for level-4 DED were: (1) severe constant discomfort of the eye and (2)

16 constant visual diminution affecting the lifestyle; (3) corneal fluorescein staining (FL)

17 score > 6 in a standardized scoring scheme22 with a maximum score of 12; (4)

18 TBUT ≤ 1 s; (5) ST ≤ 2 mm/5 min.

19 Transplantation procedures for microvascular submandibular and minor

20 salivary glands

21 SMG transplantation was performed as previously described13 (Figure 1). In brief,

6
 1 under general anesthesia, the SMG, including the branches of facial artery and vein

2 and Wharton’s duct, was harvested from the submandibular triangle and transferred

3 to the temporal region. The branches of the vessels from or going into the SMG had

4 to be harvested together. Next, the facial artery and facial veins were subjected to

5 anastomosis with the superficial temporal artery and vein, respectively. After

6 subcutaneously passing through a tunnel prepared to the upper lateral conjunctival

7 fornix, the distal end of Wharton’s duct was sutured in the upper lateral conjunctival

8 fold as an opening.

9 MSGs transplantation was performed as previously described23 (Figure 2).

10 Before surgery, the minor salivary glands flow rate (MSGFR) of three sites (upper

11 labial, lower labial, and buccal mucosa) was measured and calculated as previously

23,24
12 described . The lower or upper labial glands with higher flow rates were used as

13 the donor sites. In cases where the flow rate of upper and lower labial glands was

14 much lower than that of the buccal glands, the latter was used as a donor. Under

15 general anesthesia, the graft was obtained from the donor bed and composed of

16 salivary lobules and the covering mucosa. The recipient beds were prepared in the

17 upper and lower bulbar conjunctiva and near the fornix. The graft's mucosa was

18 covered by 8-0 Vicryl absorbable sutures and anchored to the underlying orbital

19 septum with one interrupted suture passing through the donor tissue to achieve good

20 contact between the graft and the graft underlying recipient bed. No other

21 compression methods were used.

7
 1 Postoperative treatments

2 Antibiotics were given for 4-5 days after the operation. Patients who underwent SMG

3 transplantation paid special attention to protecting the anastomosis site from being

4 pressed during the first 2 postoperative weeks. Also, capsaicin and carbachol were

5 administered for 3 months after the operation to prevent Wharton’s duct

6 obstruction.25

7 Data collection and follow-up

8 The medical records were reviewed for all patients, and they included demographic

9 features, detailed disease history, length of hospital stay (LOS), length of operation

10 (LOO), and hospital fee. The hospital fee was the total expense charged by the

11 hospital during hospitalization, which included all the medical items (e.g., the fees for

12 surgery, anesthesia, medicine, medical materials, etc.). These data were acquired

13 directly from the hospital information system.

14 Patients were followed up for 4.3, 5.8, and 3.7 years in groups A, B, and C,

15 respectively. Patient’s questionnaire and the objective ocular surface disease

16 parameters, including ST, FL, and TBUT, as well as best-corrected visual acuity

17 (BCVA) at baseline, three months post-operation, and the last time of follow-up, were

18 collected.

19 The questionnaire included two items: “My dry eye symptoms were relieved after

20 treatment” and “I am satisfied with the long-term treatment effect”. Each item was

8
 1 analyzed using a 5-point Likert-type scale, where: "completely disagree" (1 point) to

2 "completely agree" (5 points). A score of 4 or 5 was regarded as “subjective relief of

3 DED symptoms” or “satisfaction with the surgery”. Patients were asked to complete

4 the questionnaire independently. For those with poor vision (unable to read), the

5 items were read aloud by a non-related person (i.e., by someone other than the

6 doctor or the patient's relatives).

7 The patients rested 30 minutes, during which they did not engage in any kind of

8 physical activity or glandular stimulation before the ophthalmologic examination so

9 as to avoid the influence of local hyperthermia and physical activity on the secretion

10 of transplanted SMGs. The same ophthalmologist conducted the procedures

11 following the guideline of the Chinese expert consensus on clinical diagnosis and

12 treatment of dry eye22 in clinics with a temperature of 23 °C and relative humidity of

13 40%.

14 BCVA was firstly tested, followed by FL, TBUT, and the ST. A stopwatch was

15 used for timing. BCVA measurement was applied with spectacle or contact lens

16 correction. The standard logarithmic visual acuity chart (National Standard of the

17 People's Republic of China GB11533-2011) was used. Test chart background

18 luminance was ≥200 cd/m. In the FL test, the corneal surface was divided into 4

19 quadrants: upper nasal, lower nasal, upper temporal, and lower temporal, which

20 were individually scored. The fluorescein score was analyzed as follows: 0 = no

21 staining, 1 = minimal staining, 2 = mild/moderate staining, 3 = severe staining. The

9
 1 sum of the above indicators was taken as the final FL score. In the TBUT test,

2 patients were asked to blink 3 times after staining with the fluorescein strip. The time

3 from the last eye-opening to the appearance of the first dry spot was measured 3

4 times. The mean value was the TBUT score. ST was performed for 5 min using

5 Whatman No. 41 paper strips (35 × 5 mm. Tianjin Jingming New Technological

6 Development Co., Ltd) without topical anesthesia. The length of the moistened paper

7 strips was used as the score.

8 Statistical analysis

9 Considering the clinical interpretation, the baseline characteristics and longitudinal

10 clinical outcomes of group A and group C were separately compared with that of

11 group B. The comparisons between group A and group B indicated the differences in

12 surgical modalities (SMG transplantation vs. MSGs transplantation), while the

13 comparisons between groups C and B revealed the differences in DED severity

14 levels (level-2/3 DED vs. level-4 DED). For baseline characteristics, continuous

15 variables were compared between groups using independent-samples t-test, while

16 categorical variables were compared using the Chi-square test or Fisher's exact test.

17 For comparing hospital parameters and subjective long-term follow-up data (relief of

18 symptoms and overall satisfaction), general linear regression and logistic regression

19 (or exact logistic regression to deal with separation) were respectively used with

20 adjustment for age and gender to account for possible confounding bias.

21 To evaluate the objective treatment effects within each group, the values of four

10
 1 objective ocular surface disease parameters at three months post-operation and the

2 long-term follow-up (>2 years) were compared with the baseline (pre-operative)

3 values using paired-samples t-test. Differences in treatment effects between groups

4 (A vs. B, and B vs. C) were then tested using a linear mixed-effects model, with

5 objective clinical parameters as dependent variables. Random intercept per eye was

6 used to account for individual-level differences in clinical parameters. Fixed effects

7 included time variables (three time-points) and patient group, with adjustment for age

8 and gender. Interaction between time and the patient group was tested to determine

9 the between-group differences in the improvements of these clinical parameters over

10 time.

11 All analyses were conducted using SPSS 20.0 (SPSS Inc., Chicago, Illinois,

12 USA) and Stata 14.0 (StataCorp, College Station, TX, USA). P-value < 0.05 (two-

13 tailed) was considered statistically significant.

14

15 RESULTS

16 Patients

17 There were 28 patients in group A, 19 patients in group B, and 14 patients in group

18 C. All patients were diagnosed with bilateral DED. In group A and group C, 25% and

19 28.6% of patients received bilateral surgeries, respectively, compared to 5.3% in

20 group B. In total, 73 eyes were included in the analysis, and the data collections and

11
 1 analysis were based on “eye”. In addition, there was no difference in age and gender

2 between participants in group A and group B (P > 0.05), while participants in group C

3 were slightly younger. Also, there were more male patients in group C compared to

4 group B (P < 0.05; Table 1).

5 Baseline data of the disease and donors

6 The DED in all of the 35 eyes in group A was caused by SJS, which was also the

7 etiology of DED for most eyes (17 eyes, 85%) in group B. The remaining 3 (15%)

8 eyes in group B were suffering from GVHD. For most eyes (12 eyes, 66.7%) in group

9 C, the disease was caused by adenoviral conjunctivitis. While for the remaining 6

10 (33.3%) eyes in group C, the etiology of DED was not clear. The mean disease

11 duration in group A and B was more than 10 years, while in group C, it was 5.1±3.2

12 years; the difference was statistically significant (P<0.05, Table 1). Eyes from both

13 group A and B suffered the most severe damage of the lacrimal gland and ocular

14 surface, with the mean values of ST, TBUT, FL, and BCVA of 0.36±0.65, 0.03±0.17,

15 10.97±1.94, 0.11±0.14 and 0.55±1.05, 0.05±0.22, 11.10±1.65, 0.15±0.24,

16 respectively (all P values for between-group difference > 0.05). The mean values of

17 the ST, TBUT, FL, and BCVA in group C were 0.89±1.02, 3.49±1.36, 1.83±1.76, and

18 0.81±0.19, respectively, showing less severity compared to group B (P values < 0.05

19 except for ST; Table 2).

20 The donors of MSGs were harvested from similar sites in groups B and C. The

21 donor secretory functions of group C were better than in group B, as the MSGFR of

12
 1 group C was significantly higher (P < 0.05; Table 1). Although the donor sizes of

2 group C were smaller than group B, the total flow rate of the grafts (size×MSGFR)

3 was still significantly higher in group C (P < 0.05; Table 1).

4 Treatment costs and complications

5 The surgical trauma and hospital costs were significantly higher in group A and

6 similar between group B and C, as reflected by LOS, LOO, and hospital fees (Table

7 1). These results did not significantly change after adjusting for age and gender. In

8 group A, surgery was not successful for 2 eyes (5.7%), and the grafts were lost

9 because of vascular thrombosis after transplantation. Thirteen eyes (39.4%) of group

10 A developed intermittent corneal epithelial microcystic edema, and patients

11 complained of blurred vision. The hypotonic saliva elicited corneal edema when

12 postoperative epiphora occurred, which is a relatively common complication of SMG

13 transplantation.12,13 Epiphora were managed by surgical reduction of graft, topical

14 atropine gel, and botulinum toxin injection. Ranula and Wharton’s duct obstruction

15 were reported in 1 eye (3%), respectively, and were surgically cured.

16 All transplantations were successful in groups B and C. Only in 1 eye (5%) of

17 group B partial graft developed necrosis early after the operation. The residual tissue

18 showed good healing after local debridement. Partial grafts in the lower lid were

19 visible and led to cosmetic problems after the operation in 2 eyes (10%) from group

20 B, and 1 eye (5.6%) from group C. Local transient hypaesthesia of the lower lip was

21 reported in 7 eyes (35%) from group B and 5 eyes (27.8%) from group C, showing

13
 1 spontaneous remission within 6 months. The complication rates did not differ

2 between the two groups (P > 0.05).

3 Intra-group comparisons of objective parameters before and after operations

4 Objective examination results were missing for 2 eyes from group A and 1 eye from

5 group C at three-month and > 2 years post-surgery and 1 eye from group B at > 2

6 years post-surgery. The objective parameters were analyzed for 31 eyes of group A,

7 20 eyes of group B, and 17 eyes of group C at the three-month, and 31 eyes of

8 group A, 19 eyes of group B, and 17 eyes of group C at > 2 years follow-up.

9 For group A and C, the results of the ST, TBUT, and FL were all significantly

10 improved at the three-month follow-up and long-term follow-up compared with the

11 pre-operative values (all P < 0.01; Table 2). In contrast, the results of the ST, FL but

12 not TBUT, significantly improved at the two follow-ups in group B (both P < 0.05).

13 The BCVA was significantly improved in group A at the long-term follow-up (P < 0.05,

14 Table 2).

15 Between-group comparisons of objective parameters before and after

16 operations

17 To exclude the possible impact of objective baseline parameters on the follow-up

18 results, changes in objective parameters before and after operations were used in

19 between-group comparisons. A Linear mixed-effects model was used. The results

20 showed that group A had significantly larger improvement in the ST, TBUT, and FL

14
 1 at both three-month and long-term follow-up compared with group B (P for

2 time×group interaction < 0.01; Figure 3). The longitudinal changes in BCVA did not

3 significantly differ between groups A and B (P>0.05). Compared with group B, group

4 C had a significantly larger magnitude of improvements in the ST and TBUT (P for

5 time×group interaction < 0.01), but not in FL and BCVA at both follow-ups (Figure 4).

6 Between-group comparisons of patients questionnaire results

7 Patients' questionnaires were obtained for all the eyes except for the 2 eyes with

8 unsuccessful transplantation in group A. The questionnaire results were analyzed for

9 33 eyes of group A, 20 eyes of group B, and 18 eyes of group C at > 2 years follow-

10 up. Group A and C showed a higher relief rate of DED symptoms (100% and 83.3%)

11 than group B (60.0%). Results of logistic regression indicated a significant difference

12 in subjective relief rate between group A and B (P < 0.001) but not between group B

13 and C (P = 0.589) after adjusting for age and gender. Similarly, the overall subjective

14 satisfaction was significantly higher in group A (93.9%) than in group B (70.0%, P =

15 0.032); no significant difference was detected between groups B and C (83.3%, P =

16 0.892). Two group B patients who did not experience noticeable relief from the DED

17 symptoms expressed satisfaction considering they got rid of the symblepharon after

18 the operation.

19

20 DISCUSSION

15
 1 This retrospective cohort study compared the efficacies of SMG transplantation and

2 MSGs transplantation treatment in 73 eyes with different refractory DED. In the end-

3 stage DED cases with severe impairment of the eye secondary to cicatrizing

4 conjunctivitis, SMG transplantation showed a very good treatment effect (group A).

5 Abundant lubrication (a post-operative ST value of 20.23 mm/5min) provided by

6 SMG significantly improved tear film stability and ocular surface, as shown in TBUT,

7 FL, and BCVA examinations, and all patients in this group experienced the relief of

8 the DED symptoms. In contrast, in most severe DED cases, MSGs transplantation

9 led to lower lubrication (3.79 mm/5min) in group B. The TBUT and BCVA showed no

10 improvements, and the relief rate of DED symptoms was only 60%. Compared with

11 group B, patients of group C suffered from less severe DED, which was secondary to

12 non-cicatrizing conjunctivitis, and received satisfying treatment effects from MSGs

13 transplantation. The ST increased from 0.89 mm/5min to 9.35 mm/5min, TBUT

14 increased from 3.49 s to 9.08 s, FL score was reduced from 1.83 to 0.53, and 83.3%

15 of patients felt the relief of the symptoms. The between-group comparison further

16 confirmed that SMG transplantation was significantly superior to MSGs

17 transplantation for treating end-stage DED secondary to cicatrizing conjunctivitis. In

18 addition, for DED secondary to non-cicatrizing conjunctivitis with less severe eye

19 impairment, MSGs transplantation offered a significantly superior treatment effect

20 compared with end-stage DED. These conclusions were also verified by comparing

21 the rates of bilateral operations. Although all patients in the present study had

22 bilateral DED, we insisted that the operation for the other eye should be performed at

16
 1 least six months after the initial operation. Contralateral surgery provided definite

2 evidence that patients were satisfied with the treatment effect of the initial operation.

3 In the present study, 25% of patients from group A and 28.6% of patients from group

4 C underwent another surgery for the other eye after the initial operation, while this

5 was the case with only one patient from group B (5.3%).

6 The treatment costs and complications were also compared among groups.

7 SMG transplantation, which is small organ transplantation that requires vascular

8 anastomosis, showed significantly higher LOS, LOO, and hospital fees compared to

9 MSGs transplantation, which is a free tissue graft that does not require any vascular

10 anastomosis. It must be pointed out that only the major economic spending of the

11 patients (hospital fees) was included, while other expenses like travel costs were not

12 included. All MSGs transplantations were successful, while the SMG transplantation

13 was unsuccessful in two eyes. Besides, epiphora occurred in 39.4% of the eyes,

14 thus requiring operation or other management after SMG transplantation, which was

15 consistent with the literature reports.7,26 In contrast, except for one eye, there were

16 no complications requiring a secondary medical intervention after MSGs

17 transplantation. The surgical trauma, risk, and treatment burden should be taken into

18 full consideration before SMG transplantation. Accordingly, we did not perform SMG

19 transplantation for the relatively less severe DED. This is the reason why there were

20 only 3 groups of patients in this study.

21 Considering both risks and benefits, for refractory DED patients who do not have

17
 1 other treatments as an option, the surgical modalities should be chosen according to

2 the severity of the disease. For the DED secondary to non-cicatrizing conjunctivitis

3 and those with less severe impairment of the tear film stability and ocular surface

4 (e.g., group C), MSGs transplantation might be recommended as a first choice. Most

5 patients could benefit from adequate lubrication and substantial improvements with

6 minor treatment risk and cost. However, for the end-stage DED secondary to

7 cicatrizing conjunctivitis (as in group A or B of this study), the treatment effect of

8 MSGs transplantation may be limited (less than 4 mm/5min of lubrication and 60% of

9 symptoms relief rate); thus, SMG transplantation might be recommended.

10 The differences observed in the treatment effect of MSGs transplantation on

11 different kinds of DED may be explained as follows: firstly, most DEDs in group B

12 were caused by SJS, which could impair not only the lacrimal gland but also the

13 MSGs.27 In contrast, adenoviral conjunctivitis does not have an impact on MSGs.

14 Our data of donor secretory functions at baseline confirmed the significantly higher

15 secretory flow rate of MSGs in group C than in group B. A previous study suggested

16 that the preoperative flow rate of MSGs is positively correlated with postoperative

17 lubrication and the treatment effect.23 Consequently, group C had better treatment

18 results than group B. Secondly, cicatrizing conjunctivitis such as SJS causes severe

19 scar formation in the suffered eye, which is the location of the recipient bed of the

20 free grafted MSGs tissues. The poor condition of the recipient bed is likely to be

21 harmful to the survival of the grafted tissues. At the same time, this pitfall was not

22 found in patients with adenoviral conjunctivitis.

18
 1 The different treatment effects of SMG transplantation and MSGs transplantation

2 on the most severe DED may be explained as follows: as a major salivary gland,

3 SMG has a much stronger secretory function compared with MSGs. Besides, SMG

4 transplantation could preserve the function to the greatest extent considering that the

5 blood circulation is rebuilt during operation. Thus, the amount of lubrication after

6 SMG transplantation is much greater than that after MSGs transplantation.

7 Considering that most of the patients suffered from bilateral DEDs, the

8 proportion of patients undergoing bilateral SMG transplantations was relatively

9 limited, which was consisted with the reports from other groups 4,5,6,7,9,10,11. We got a

10 possible explanation for the cause of this issue based on our communications with

11 the patients. These patients with end-stage DED experienced a marked decrease in

12 vision and the fears of blindness and disability were the most powerful motivation for

13 accepting the surgery. After the one-side operation, the eye vision might be

14 preserved, and the patients can already be protected from being blind and disabled.

15 They might then lose the motivation to accept one more time of organ transplantation

16 surgery. However, further evidence is needed for this explanation.

17 This study has a few limitations. First, as a retrospective cohort study, a grouping

18 of the patients was not random; future randomized controlled trials of surgical

19 modalities are warranted to validate our findings. Moreover, only 3 study groups

20 were examined; we did not perform SMG transplantation for the relatively less

21 severe DED. However, a negative correlation between subjective symptoms and

19
 1 objective examination is not rare in DED.28 Patients could have severe pain with only

2 relatively moderate impairment of the eye. It is still unclear whether these patients

3 could benefit from SMG transplantation when other treatments failed, and MSGs

4 transplantation does not provide an effective cure.

6 Funding: This work was supported by the National Natural Science Foundation of

7 China [82170977, 81974151] and PKUSS fund [PKUSSNCT-20A01].

9 Financial Disclosures: No financial disclosures.

10

11 Other Acknowledgments: NA.

20
References

1. Jones L, Downie LE, Korb D, et al. TFOS DEWS II management and

therapy report. Ocul Surf 2017; 15(3): 575-628.

2. Murube J. Transplantation of salivary gland to the lacrimal basin. Scand J

Rheumatol Suppl 1986; 61: 264-7.

3. Kumar PA, Hickey MJ, Gurusinghe CJ, O'Brien BM. Long term results of
submandibular gland transfer for the management of xerophthalmia. Br J
Plast Surg 1991; 44(7): 506-8.

4. MacLeod AM, Robbins SP. Submandibular gland transfer in the correction

of dry eye. Aust N Z J Ophthalmol 1992; 20(2): 99-103.

5. Geerling G, Sieg P, Bastian GO, Laqua H. Transplantation of the

autologous submandibular gland for most severe cases of
keratoconjunctivitis sicca. Ophthalmology 1998; 105(2): 327-35.

6. Jia G, Wang Y, Lu L, Wang X, Li Z. Reconstructive lacrimal gland with free

submandibular gland transfer for management of xerophthalmia.
Zhonghua Yan Ke Za Zhi 1998; 34(5): 388-90. (In Chinese)

7. Sieg P, Geerling G, Kosmehl H, Lauer I, Warnecke K, von Domarus H.

Microvascular submandibular gland transfer for severe cases of
keratoconjunctivitis sicca. Plast Reconstr Surg 2000; 106(3): 554-60;
Discussion 61-2.

8. Yu GY, Zhu ZH, Mao C, et al. Microvascular autologous submandibular

gland transfer in severe cases of keratoconjunctivitis sicca. Int J Oral
Maxillofac Surg 2004; 33(3): 235-9.

9. Paniello RC. Submandibular gland transfer for severe xerophthalmia.

Laryngoscope 2007; 117(1): 40-4.

10. Jacobsen HC, Hakim SG, Lauer I, Dendorfer A, Wedel T, Sieg P. Long-
term results of autologous submandibular gland transfer for the surgical
treatment of severe keratoconjunctivitis sicca. J Craniomaxillofac Surg
2008; 36(4): 227-33.

11. Borrelli M, Schroder C, Dart JK, et al. Long-term follow-up after

submandibular gland transplantation in severe dry eyes secondary to
cicatrizing conjunctivitis. Am J Ophthalmol 2010; 150(6): 894-904.

12. Su JZ, Zheng B, Liu XJ, et al. Quality of life and patient satisfaction after
submandibular gland transplantation in patients with severe dry eye

21
 disease. Ocul Surf 2019; 17(3): 470-5.

13. Zhang L, Su JZ, Cai ZG, et al. Factors influencing the long-term results of
autologous microvascular submandibular gland transplantation for severe
dry eye disease. Int J Oral Maxillofac Surg 2019; 48(1): 40-7.

14. Murube J, Manyari A, ChenZhuo L. Labial salivary gland transplantation in

severe dry eye. Oculoplast Orbital Reconstr Surg 1998; 1: 104-10.

15. Guerrissi JO, Belmonte J. Surgical treatment of dry eye syndrome:

conjunctival graft of the minor salivary gland. J Craniofac Surg 2004;
15(1): 6-10.

16. Soares EJ, Franca VP. Transplantation of labial salivary glands for severe
dry eye treatment. Arq Bras Oftalmol 2005; 68(4): 481-9.

17. Geerling G, Raus P, Murube J. Minor salivary gland transplantation. Dev

Ophthalmol 2008; 41: 243-54.

18. Marinho DR, Burmann TG, Kwitko S. Labial salivary gland transplantation
for severe dry eye due to chemical burns and Stevens-Johnson
syndrome. Ophthalmic Plast Reconstr Surg 2010; 26(3): 182-4.

19. Sant' Anna AE, Hazarbassanov RM, de Freitas D, Gomes JA. Minor
salivary glands and labial mucous membrane graft in the treatment of
severe symblepharon and dry eye in patients with Stevens-Johnson
syndrome. Br J Ophthalmol 2012; 96(2): 234-9.

20. Wakamatsu TH, Sant'Anna A, Cristovam PC, Alves VAF, Wakamatsu A,

Gomes JAP. Minor salivary gland transplantation for severe dry eyes.
Cornea 2017; 36(p): S26-S33.

21. Lemp MA, Baudouin C, Baum J, et al. The definition and classification of
dry eye disease: report of the definition and classification subcommittee of
the International Dry Eye WorkShop (2007). Ocul Surf 2007; 5(2): 75-92.

22. Chinese Medical Association OB. The keratonosus group: expert

consensus on clinical diagnosis and treatment of dry eye. Chin J
Ophthalmol. 2020; 56(10): 741–7. (In Chinese)

23. Su JZ, Wang Z, Liu XJ, Lv L, Yu GY. Use of saliva flow rate measurement
in minor salivary glands autotransplantation for treatment of severe dry
eye disease. Br J Ophthalmol 2021. DOI: 10.1136/bjophthalmol-2020-
317552

24. Wang Z, Shen MM, Liu XJ, Si Y, Yu GY. Characteristics of the saliva flow
rates of minor salivary glands in healthy people. Arch Oral Biol 2015;

22
 60(3):385-92.

25. Su JZ, Liu XJ, Wang Y, et al. Effects of Capsaicin and carbachol on
secretion from transplanted submandibular glands and prevention of duct
obstruction. Cornea 2016; 35(4): 494-500.

26. Geerling G, Garrett JR, Paterson KL, et al. Innervation and secretory
function of transplanted human submandibular salivary glands.
Transplantation 2008; 85(1): 135-40.

27. Wang Z, Li W, Hong X, et al. Minor salivary gland function is decreased in

hyposalivation-related diseases. Arch Oral Biol 2016; 69: 63-70.

28. Pflugfelder SC, Geerling G, Kinoshita S, et al. Management and therapy

of dry eye disease: report of the management and therapy subcommittee
of the International Dry Eye WorkShop (2007). Ocul Surf 2007; 5(2): 163-
78.

23
Figure Legends

Figure 1. Procedures and treatment effect of submandibular gland

transplantation. (A) Bilateral dry eye disease secondary to Stevens-Johnson

syndrome. (B) The incision in the temporal region and dissection of the

superficial temporal vessels (left side). (C) Dissection of the submandibular

gland. (D) Dissection of the Wharton’s duct. (E) The donor, including the

submandibular gland, the facial vessels, and Wharton’s duct. (F) Anastomosis

of the vessels (arrows). (G) Reopening of the Wharton’s duct in the eye

(arrow). A nylon tube was inserted and left in the duct for 7 days. (H) Follow-

up image 9 years after the operation. Compared with the untreated right eye,

the left eye had plenty of lubrication and better ocular surface condition.

Figure 2. Procedures and treatment effect of minor salivary glands

transplantation. (A) Bilateral dry eye disease secondary to adenoviral

conjunctivitis. (B) The grafts were harvested from the lower lip above the

muscles, and branches of the trigeminal nerve (arrows) were preserved. (C)

Two pieces of salivary lobules with the covering mucosa. (D) The grafts were

transplanted and fixed in the left eye. (E) The wounds of the lip were repaired

with an acellular dermal matrix. (F, G) The right eye was treated 6 months

later with the grafts from the upper lip. (H-J) Follow-up images 4 years after

the second operation. The incisions of the lips healed well, and the dry eye

symptoms were relieved in both eyes.

24
 Figure 3. Longitudinal changes of objective clinical parameters in group

A compared to group B. Group A had significantly larger improvements in

the Schirmer test, tear break-up time, and corneal fluorescein staining at

three-month and > 2 years follow-up compared with group B (P < 0.01). (A)

Schirmer test. (B) Tear break-up time. (C) Corneal fluorescein staining. (D)

Best-corrected visual acuity.

Figure 4. Longitudinal changes of objective clinical parameters in group

B compared to group C. Group C had significantly larger improvements in

the Schirmer test and Tear break-up time at three-month and > 2 years follow-

up compared with group B (P < 0.01). (A) Schirmer test. (B) Tear break-up

time. (C) Corneal fluorescein staining. (D) Best-corrected visual acuity.

Table 1. Baseline data of three groups

P-value
Characteristics Group A Group B Group C B vs.
A vs. B
C
Cicatrizing Cicatrizing Non-cicatrizing
DED etiology
conjunctivitis conjunctivitis conjunctivitis
Grouping
DED severity level level-4 level-4 < level-4
methods
SMG MSGs MSGs
Surgical modality
transplantation transplantation transplantation
Patients 28 19 14
Eyes 35 20 18
Numbers
Patients with
7 (25%) 1 (5.3%) 4 (28.6%) 0.119 0.138
bilateral surgeries
Age (year) 29.6±15.1 34.4±10.5 28.2±4.2 0.211 0.022
Demographic
Gender
features 23/12 12/8 4/14 0.672 0.019
(female/male)
SJS: 17 AC: 13
Detailed etiology SJS: 35
Disease GVHD: 3 unknown: 5
history Disease duration
11.3±11.6 10.5±9.1 5.1±3.2 0.781 0.020
(year)
Hospital LOS (day) 13.0±1.1 7.8±0.7 7.5±0.7 <0.001 0.287

25
parameters LOO (hour) 6.1±0.6 1.9±0.3 1.8±0.3 <0.001 0.310
28486.2±2740.
Hospital fee (RMB) 16006.7±2217.6 16845.6±3617.6 <0.001 0.389
0
low lip: 14 low lip: 13
Donor sites —— upper lip: 5 upper lip: 5 0.627
Donor buccal: 1 buccal: 0
parameters 2
Sizes (cm ) —— 8.8±2.5 6.6±1.6 0.003
(MSGs)
MSGFR (mg/min) —— 1.8±0.6 3.2±1.1 <0.001
Size×MSGFR —— 15.9±6.5 21.1±8.2 0.045
DED: dry eye disease; SMG: submandibular gland; MSGs: minor salivary glands; SJS:
Stevens-Johnson syndrome; GVHD: graft versus host disease; AC: acute conjunctivitis; LOS:
length of hospital stay; LOO: length of operation; MSGFR: salivary flow rate of minor salivary
glands. RMB: Renminbi

P values for group A vs. B and B vs. C were based on the Chi-square test for categorical
variables (except for Fisher's exact test for Bilateral surgery) and the independent-samples t-
test for continuous variables.

Table 2. Mean levels of objective clinical parameters of 3 groups at baseline and follow-up
Timepoints

Objective parameter Pre-operative Three-month post- Long-term (>2 years)

(71 eyes) operation follow-up
(68 eyes) (67 eyes)
Schirmer test (mm/5
min)
** **
Group A 0.36±0.65 20.29±9.41 20.23±7.31
** **
Group B 0.55±1.05 3.30±2.47 3.79±2.99
** **
Group C 0.89±1.02 7.71±4.09 9.35±7.78

TBUT (s)
** **
Group A 0.03±0.17 1.58±2.03 1.74±2.21
Group B 0.05±0.22 0.10±0.45 0.00±0.00
** **
Group C 3.49±1.36 9.53±5.68 9.08±6.26

26
 FL
** **
Group A 10.97±1.94 7.55±2.23 7.58±2.36
* *
Group B 11.10±1.65 10.10±1.77 9.58±2.17
* **
Group C 1.83±1.76 0.71±1.16* 0.53±1.33
BCVA
**
Group A 0.11±0.14 0.15±0.15 0.20±0.21
Group B 0.15±0.24 0.19±0.25 0.20±0.25
Group C 0.81±0.19 0.88±0.14 0.89±0.14
TBUT: Tear break-up time; FL: Corneal fluorescein staining. BCVA: Best-corrected visual
acuity.

* P < 0.05, ** P < 0.01, based on paired-samples t-test comparing post-surgical time-points
with pre-surgical level within each group separately. The mean duration of the long-term
follow-up were 3.2, 3.8, and 3.6 years for group A, B, and C, respectively (P > 0.10).

27