Highlights
-
•
Systemic immunosuppressant cessation/tapering is a risk factor for MGD post-HSCT.
-
•
Improved chronic ocular GVHD dose not linked to meibomian gland recovery.
-
•
Immunosuppressant adjustment necessitates meibomian gland monitoring.
Purpose
To observe the longitudinal changes in meibomian gland structure and function in patients after allogeneic hematopoietic stem cell transplantation (allo-HSCT) and to investigate the potential associated factors.
Design
Single-center retrospective clinical cohort study.
Participants
Sixty-eight patients after allo-HSCT, including 48 patients (92 eyes) with and 20 patients (40 eyes) without chronic ocular graft-versus-host disease (coGVHD).
Methods
Medical records were retrospectively reviewed. Data collected included demographic characteristics, medical history, ocular surface disease index questionnaire, corneal fluorescein staining, bulbar conjunctival injection, Schirmer’s test, noninvasive break-up time, lid margin abnormalities, meibum quality scores, and meibomian gland area ratio (MGAR). Follow-up data were compared. A generalized estimating equation was used to analyze the associated factors.
Main Outcome Measures
Change in MGAR.
Results
All patients were followed up for 22.9 ± 10.7 months since first visit (coGVHD group vs non-coGVHD group = 23.6 ± 10.8 months vs 21.3 ± 10.5 months, P =.441). Compared with baseline, the severity of coGVHD at the last follow-up visit improved (International Chronic Ocular Graft-versus-Host Disease [ICCGVHD] score: 7.3 ± 2.3 vs 8.5 ± 1.7, P <.001), but no significant change in MGAR (51.7% ± 14.8% vs 52.6% ± 14.8%, P =.341). Specifically, MGAR remained stable in 59 eyes (64.1%), deteriorated in 14 eyes (15.2%), and improved in 19 eyes (20.7%). In the non-coGVHD group, while ICCGVHD scores remained stable compared to baseline, MGAR showed a decrease at the final follow-up (70.2 ± 11.0% vs 73.5 ± 10.0%, P =.001). Cessation or tapering of systemic immunosuppressants was associated with MGAR worsening in both the coGHVD group (odds ratio [OR]: 4.60, 95% confidence interval [CI]: 1.24-17.00) and the non-coGHVD group (OR: 2.48, 95%CI: 1.14-5.40). Further subgroup analysis revealed that in coGVHD patients, the subgroup with cessation or tapering of systemic immunosuppressants showed a significant worsening in meibum quality (8.4 ± 5.3 vs 5.9 ± 4.6, P =.005).
Conclusions
Improvement in coGVHD is not necessarily accompanied by improvement in Meibomian gland dysfunction (MGD). Cessation or tapering of systemic immunosuppressants was associated with the progression of MGD after allo-HSCT.
INTRODUCTION
C hronic ocular graft vs host disease (coGVHD) is the most common ocular complication following allogeneic hematopoietic stem cell transplantation (allo-HSCT), primarily manifesting as refractory dry eye that severely impacts patients’ quality of life. ,, Meibomian gland dysfunction (MGD), a leading cause of evaporative dry eye, is considered the second most frequent ocular complication post-allo-HSCT and is highly prevalent among coGVHD patients. , Increasing evidence, including our own previous work, indicates that MGD can occur before or after HSCT regardless of the presence of coGVHD. ,,,, Current evidence suggests a complex pathogenesis underlying MGD, potentially involving factors such as underlying primary disease, chemotherapy, radiotherapy, and coGVHD.
There remains a scarcity of longitudinal cohort studies investigating changes in meibomian gland structure and function in allo-HSCT patients. To date, one cohort study reported the alteration in meibomian gland loss in patients with coGVHD. That study revealed that despite standardized treatment, 75.7% of affected eyes showed no significant improvement in meibomian gland structure or function after 1 year of comprehensive therapy, while 18.9% exhibited rapidly progressive gland dropout. In addition, 2 other cohort studies have investigated meibomian gland changes following allo-HSCT. , Both reported worsening meibomian gland loss and shortened tear break-up time in patients after HSCT, irrespective of coGVHD occurrence. Myeloablative conditioning was identified as a factor associated with meibomian gland loss progression, whereas the source of hematopoietic stem cells, donor type, sex mismatch, and pre-transplant treatments showed no significant association. Furthermore, our cross-sectional study demonstrated no significant difference in the prevalence or severity of MGD between coGVHD and non-coGVHD patients after allo-HSCT, and the severity of MGD was not correlated with the severity of coGVHD. Collectively, these findings suggest that meibomian gland alterations after allo-HSCT are linked to systemic factors.
Systemic chronic graft-vs-host disease (cGVHD) and the associated immunosuppressive regimens may influence the onset and progression of coGVHD. ,,, One study has reported early improvement in meibomian gland morphology after allo-HSCT, which the authors speculated might be related to the use of systemic immunosuppressants. Our previous research also found that patients receiving oral immunosuppressants tended to have a thicker tear film lipid layer. However, whether systemic immunosuppressants affect the progression of MGD after HSCT remains unclear and warrants further investigation. Therefore, we conducted an intermediate-term retrospective clinical cohort study focusing on coGVHD and non-coGVHD patients after allo-HSCT to identify key factors associated with MGD progression and to evaluate the impact of anti-coGVHD treatment on MGD.
METHODS
Study Design and Subjects
This single-center retrospective clinical cohort study was performed at the Zhongshan Ophthalmic Center. Demographic and clinical data of patients who underwent allo-HSCT were systematically collected through hospital medical records. The study was approved by the Ethics Committee of Zhongshan Ophthalmic Center (Approval Number: 2019KYPJ135) and was conducted in accordance with the principles of the Declaration of Helsinki.
The study enrolled adult patients who had undergone allo-HSCT, including both non-coGVHD patients and coGVHD patients who received standard treatment.
Exclusion criteria were as follows: (1) presence of other ocular diseases except for refractive errors < ± 3 dioptres or early age-related cataract; (2) history of ocular surgery, including procedures that may affect meibomian gland function, trauma, or contact lens wear within 3 months before the initial visit or during the follow-up period; (3) systemic diseases that may affect the ocular surface, such as diabetes mellitus, thyroid disorders, or sex hormone abnormalities; (4) recurrence of hematologic disease, systemic infection, corneal perforation, or ocular infection during follow-up; (5) incomplete medical records; (6) adjustment of systemic immunosuppressant therapy within 1 month before the last follow-up; and (7) development of coGVHD during follow-up among initially non-coGVHD patients.
The diagnosis and grading of coGVHD were based on the International Chronic Ocular Graft-vs-Host disease (ICCGVHD) diagnostic criteria. MGD was diagnosed according to the 2011 International Workshop on Meibomian Gland Dysfunction. The intensity of immunosuppression after HSCT was defined as follows: low intensity: prednisone < 0.5 mg/kg/d; moderate intensity: prednisone ≥ 0.5 mg/kg/d and/or any other immunosuppressive agent; high intensity: combination therapy with 2 or more agents, including prednisone ≥ 0.5 mg/kg/d.
All coGVHD patients received standard, indication-based anti-coGVHD therapy, which included one or more of the following: artificial tears, topical corticosteroid eye drops, cyclosporine A or tacrolimus eye drops, autologous serum eye drops, punctal occlusion, eyelid warm compresses, and eyelid hygiene. Patients in the non-coGVHD group received artificial tears and symptomatic management for MGD if they reported dry eye or MGD-related symptoms. Patients were followed up at 1 to 3-month intervals, adjusted according to their clinical status.
Systemic condition, including systemic cGVHD, was evaluated by hematologists, who also prescribed and adjusted the systemic immunosuppressant regimen (including agents such as corticosteroids, cyclosporine, sirolimus, mycophenolate mofetil, or ruxolitinib).
Data collection and main outcome measures
Collected data included detailed demographic information, history of hematopoietic malignancies, allo-HSCT history, and clinical characteristics. The Ocular Surface Disease Index (OSDI) questionnaire, a validated 12-item instrument scored from 0 to 100, was used to rapidly assess ocular irritation symptoms consistent with dry eye disease and their impact on visual function. The bulbar conjunctival injection (BCI) grading score was applied to evaluate the severity of conjunctival hyperemia. The non-anesthetized Schirmer I test (ST) was used to assess lacrimal gland secretion function. Corneal fluorescein staining (CFS), scored according to the National Eye Institute grading system, was used to evaluate the degree of corneal damage. Meibomian gland evaluation included the meibum quality score (MQS), lid margin abnormality score, and meibography.
The meibomian gland area ratio (MGAR) served as the main outcome measure, providing a quantitative assessment of meibomian gland loss. MGAR was calculated by acquiring meibography images and performing masked analysis with ImageJ software to determine meibomian gland dropout. Changes in meibomian gland loss were defined as follows: a decrease in meiboscore by ≥ 1 point was considered improvement, no change was defined as stable, and an increase by ≥ 1 point was defined as progression.
Statistical analysis
Data from both eyes were included in the statistical analysis. Statistical analysis was performed using IBM SPSS Statistics, version 29.0 (IBM Corp.). Continuous variables were expressed as mean ± SD or median (IQR), and categorical variables were presented as frequency (percentage). Changes in parameters were analyzed using an empty generalized estimating equation (GEE) model with the baseline-to-final paired difference as the dependent variable. To identify relevant risk factors, a GEE model was first applied, with MGD progression (dichotomized as presence or absence of progression) as the dependent variable. Independent variables included in this initial analysis were age, sex, time between HSCT and coGVHD, primary disease, source of HSCT, cessation or tapering of systemic immunosuppressants, change in the number of organs involved by GVHD, and changes in ocular surface parameters—including the ICCGVHD score, non-invasive breakup time (NIBUT), subtarsal fibrosis, and lid margin abnormality. All variables with P <.2 in the univariable analysis were subsequently entered into the multivariable GEE model to identify independent risk factors. All analyses were adjusted for inter-eye correlation. For patients with stable systemic conditions, correlations between changes in meibomian gland parameters and changes in different organ cGVHD were further analyzed: Pearson correlation for normally distributed variables, Spearman correlation for non-normally distributed or categorical variables, and Kendall’s tau-b correlation for ordered categorical variables. All statistical tests were 2-sided, and a P <.05 was considered statistically significant.
RESULTS
Subjects
The baseline characteristics of the participants are shown in Table 1 . A total of 48 patients (92 eyes) were included in the coGVHD group and 20 patients (40 eyes) in the non-coGVHD group. No significant differences were found in age or sex between the groups. In the coGVHD group, the mean age was 35.1 ± 9.4 years (range, 19- 60 years), and 68.8% were male. In the non-coGVHD group, the mean age was 32.8 ± 12.2 years (range, 20- 62 years), and 50% were male.
Table 1
Demographic and Transplant Characteristics of Patients After Allogeneic Hematopoietic Stem Cell Transplantation.
|
coGVHD
( N = 48 Patients) |
non-coGVHD
( N = 20 Patients) |
|
|---|---|---|
|
Age, years, mean ± SD,
Range |
35.1 ± 9.4,
19– 60 |
32.8 ± 12.2,
20– 62 |
| Male/female, n | 33/15 | 10/10 |
|
Mean follow-up time, months, mean ± SD,
Range |
23.6 ± 10.8
6.9, 44.6 |
21.3 ± 10.5
6.4, 39.1 |
| Time between HSCT and first visit, months, median, | 24.3 | 10.3 |
| IQR | 14.9, 43.5 | 6.0, 40.1 |
| Primary disease, n (%) | ||
| AML | 24 (50.0) | 12 (60.0) |
| ALL | 14 (29.2) | 4 (20.0) |
| CML | 2 (4.2) | 0 (0) |
| MDS | 4 (8.3) | 2 (10.0) |
| Others | 4 (8.3) | 2 (10.0) |
| Radiotherapy, n (%) | 10 (20.8) | 4 (20.0) |
| Chemotherapy, n (%) | 47 (97.9) | 20 (100.0) |
| Source of HSCT, n (%) | ||
| PBT | 23 (47.9) | 3 (15.0) |
| BMT | 19 (39.6) | 7 (35.0) |
| PBT + BMT | 6 (12.5) | 8 (40.0) |
| PBT + BMT + UCBT | 0 | 1 (5.0) |
| PBT + UCBT | 0 | 1 (5.0) |
| Donor type, n (%) | ||
| MRD | 36 (75.0) | 6 (30.0) |
| MMRD | 8 (16.7) | 12 (60.0) |
| MUD | 3 (6.3) | 1 (5.0) |
| MMUD | 1 (2.1) | 1 (5.0) |
| Systemic cGVHD, n (%) | 45 (93.8) | 4 (20.0) |
| Systemic use of immunosuppressants, n (%) | 34 (70.8) | 13 (65.0) |
| Cyclosporine | 19 (39.6) | 9 (45.0) |
| Corticosteroids | 16 (33.3) | 4 (20.0) |
| Ruxolitinib | 9 (18.8) | 3 (15.0) |
| Tacrolimus | 7 (14.6) | 3 (15.0) |
| Mycophenolate mofetil | 6 (12.5) | 1 (5.0) |
| Sirolimus | 3 (6.3) | 0 |
| Rituximab | 1 (2.1) | 0 |
| Severity of coGVHD, n (%) | 0 | |
| Mild/moderate | 19 (39.6) | |
| Severe | 29 (60.4) |
coGVHD, chronic ocular graft-versus-host disease; SD, standard deviation; HSCT, hematopoietic stem cell transplantation; IQR, interquartile range; AML, acute myeloid leukemia; ALL, acute lymphocytic leukemia; CML, chromic myeloid leukemia; MDS, myelodysplastic syndrome; PBT, peripheral blood transplantation; BMT, bone marrow transplantation; UCBT, umbilical cord blood transplantation; MRD, matched related donor; MMRD, mismatched related donor; MUD, matched unrelated donor; MMUD, mismatched unrelated donor; cGVHD, chronic graft-versus-host disease.
The 2 groups were followed for a contemporaneous period since the first visit (coGVHD: 23.6 ± 10.8 months, non-coGVHD: 21.3 ± 10.5 months, P =.441). In the coGVHD group, the median time from allo-HSCT to the first visit was 24.3 months (14.9, 43.5), while the corresponding time for the non-coGVHD group was 10.3 months (6.0, 40.1). Acute myeloid leukemia was the most common primary hematologic disease in both groups. The predominant stem cell source in the coGVHD patients was peripheral blood (23/48, 47.9%), whereas in the non-coGVHD patients, it was combined peripheral blood and bone marrow transplantation (8/20, 40%). Forty-seven out of 48 coGVHD patients and all non-coGVHD patients had received chemotherapy before allo-HSCT. Radiotherapy before allo-HSCT was administered in 10 of 48 coGVHD patients (20.8%) and 4 of 20 non-coGVHD patients (20.0%).
Systemic GVHD was present in 93.8% (45/48) of patients in the coGVHD group, and 70.8% were on systemic immunosuppressant therapy at the initial visit. By contrast, only 20% (4/20) of patients in the non-coGVHD group had systemic GVHD, and 65% were receiving systemic immunosuppressant. According to the ICCGVHD criteria, 60.4% (29/48) of patients in the coGVHD group had severe coGVHD at their initial visit.
No significant improvement in MGAR despite clinical improvement in coGVHD patients; progression of MGAR in non-coGVHD patients
Changes in clinical and MGD parameters before and after treatment are presented in Tables 2 and 3 . In the coGVHD group, the severity of coGVHD was significantly improved compared with baseline (ICCGVHD score: 7.3 ± 2.3 vs 8.5 ± 1.7, P <.001), whereas no significant change was observed in the non-coGVHD group (ICCGVHD score: 2.5 ± 1.5 vs 2.4 ± 1.9, P = .823). Specifically, at last visit, significant improvements compared to baseline were noted in the coGVHD group in OSDI score (32.0 ± 23.4 vs 47.5 ± 21.7, P <.001), CFS score (7.9 ± 4.1 vs 9.6 ± 4.6, P =.016), and BCI grade (0.8 ± 0.6 vs 1.0 ± 0.7, P <.001). In contrast, the non-coGVHD patients exhibited no significant change in those parameters.
Table 2
Ocular Surface and MGD Changes By Systemic Immunosuppression Adjustment Status in coGVHD Patients.
|
Total
( n = 92 eyes) |
Cessation/Tapering Systemic Immunosuppressants ( n = 61 eyes) | No Cessation/Tapering systemic Immunosuppressants ( n = 31 Eyes) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| V 0 | V L | P | V 0 | V L | P | V 0 | V L | P | |
| ICCGVHD score | 8.5 ± 1.7 | 7.3 ± 2.3 | <.001 | 8.7 ± 1.7 | 7.4 ± 2.2 | <.001 | 8.1 ± 1.7 | 7.0 ± 2.3 | .004 |
| OSDI | 47.5 ± 21.7 | 32.0 ± 23.4 | <.001 | 50.4 ± 18.6 | 33.3 ± 24.2 | <.001 | 41.8 ± 26.2 | 29.2 ± 21.7 | .045 |
| ST (mm/5 min) | 3.4 ± 3.0 | 3.2 ± 3.5 | .755 | 3.5 ± 2.9 | 2.9 ± 3.9 | .239 | 3.0 ± 3.2 | 3.8 ± 2.4 | .310 |
| CFS | 9.6 ± 4.6 | 7.9 ± 4.1 | .016 | 9.1 ± 4.7 | 8.3 ± 4.2 | .323 | 10.6 ± 4.5 | 7.3 ± 4.0 | .005 |
| BCI | 1.0 ± 0.7 | 0.8 ± 0.6 | <.001 | 1.1 ± 0.7 | 0.8 ± 0.6 | <.001 | 0.8 ± 0.7 | 0.7 ± 0.5 | .637 |
| NIBUT (s) | 3.8 ± 3.5 | 3.2 ± 3.1 | .286 | 3.9 ± 3.3 | 3.5 ± 3.5 | .576 | 3.6 ± 4.0 | 2.6 ± 2.3 | .276 |
| MQS | 6.2 ± 4.8 | 7.7 ± 5.2 | .036 | 5.9 ± 4.6 | 8.4 ± 5.3 | .005 | 6.7 ± 5.2 | 6.4 ± 4.7 | .803 |
| MGAR (%) | 52.6 ± 14.8 | 51.7 ± 14.8 | .341 | 50.6 ± 15.7 | 49.3 ± 15.8 | .346 | 56.5 ± 12.2 | 56.3 ± 11.6 | .846 |
| Upper lid | 55.8 ± 16.6 | 54.1 ± 16.7 | .163 | 53.4 ± 18.3 | 51.5 ± 18.7 | .253 | 60.6 ± 11.6 | 59.3 ± 10.0 | .399 |
| Lower lid | 48.4 ± 17.2 | 48.9 ± 16.2 | .784 | 46.7 ± 17.0 | 47.1 ± 16.3 | .860 | 51.8 ± 17.4 | 52.5 ± 15.7 | .694 |
| Lid margin abnormality | 2.7 ± 1.2 | 2.7 ± 1.2 | .934 | 2.6 ± 1.2 | 2.7 ± 1.3 | .488 | 2.9 ± 1.1 | 2.7 ± 1.1 | .340 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
