For various hematologic malignancies and severe immunodeficiencies, allogeneic hematopoietic stem cell transplantation (allo-HSCT) is a curative treatment modality (1). Annually, approximately 30,000 allo-HSCT procedures are conducted worldwide, with a consistent upward trend in transplant numbers (1, 2). Graft-versus-host disease (GVHD), which is the major complication of allo-HSCT, can affect many tissues including the skin, liver, gut and eye, increasing the risk of morbidity and mortality in the post-treatment period (3, 4). GVHD is a multi-organ disease derived from immune dysregulation and tissue inflammation with single or multisystem involvement, resulting in tissue fibrosis and organ dysfunction (5). GVHD can occur in 10~80% of transplant recipients with ocular involvement observed in 40~60% of cases (6, 7), which indicates that oGVHD is more common in patients after transplantation and has certain indicative significance for rejection after transplantation. GVHD involving the eyes is generally a chronic disease. Chronic graft-versus-host disease-related dry eye (cGVHD-DE) is the most common manifestation of oGVHD (8), which is manifested as dry eye, foreign body sensation, severe light sensitivity, chronic conjunctivitis, periorbital hyperpigmentation, itching and eye tingling after allo-HSCT, resulting in obvious eye discomfort, decreased vision, and even blindness (9–11).

At present, the widely recognized diagnostic criteria for oGVHD include The National Institutes of Health Consensus Conference (NIH CC) 2014 criteria (9) and The International Chronic oGVHD (ICCGVHD) consensus group diagnostic criteria (12). A comparative study of the NIH 2014 criteria and ICCGVHD criteria found that the two have a moderate agreement, but the ICCGVHD criteria were noted to be better at differentiating oGVHD patients from non‐oGVHD dry eye disease (DED), due to its more stringent criteria which also considers the status of systemic GVHD (13). The diagnosis of chronic oGVHD(coGVHD) is mainly based on the presence of ocular manifestations such as dry eye and ocular surface damage. However, the feature of dry eye in oGVHD overlaps those of dry eye disease though they have distinct etiologies, presentation, pathophysiology, clinical manifestations, and treatments (14). Tears are secreted by the lacrimal gland, and a reduction in tear production can precede the onset of dry eye. Consequently, changes in the morphology of the lacrimal gland may occur before the manifestation of dry eye. As the disease progresses to advanced stages, suffering a significant decline in the quality of life and limited effective treatment options potentially impose a substantial financial burden on their families. Therefore, it is of great significance and importance to find more reasonable and integrated methods for early and objective diagnosis of oGVHD (13).

Another condition characterized by dry eye as its primary clinical manifestation is Sjogren’s disease (SjD), a chronic autoimmune connective tissue disease presenting a triad of symptoms including sicca symptoms, fatigue, and pain (15). Sjogren’s disease is histopathologically characterized by lymphocytic infiltration of exocrine glands including lacrimal leading to dry eyes(keratoconjunctivitis sicca) (16, 17). Studies using mouse models of chronic GVHD have revealed inflammatory changes in lacrimal glands, featuring ductal epithelia inflammatory cell infiltration, including eosinophils, macrophages, CD8⁺T cells together with some CD4⁺T cells and finally leading to fibrosis around the lacrimal gland ducts (18, 19). A recent research observed a reduced proportion of epithelial cell populations and different gene expressions in GVHD lacrimal glands compared with non-GVHD, strengthening the relationship between the lacrimal gland and the development of oGVHD (20). Interestingly, the histological characteristics of lacrimal gland involvement in Sjogren’s disease and oGVHD appear similar. Ultrasonography (US) is widely used in the diagnosis of numerous diseases due to its non-invasive nature, ease of use, widespread availability, and real-time assessment capabilities. Salivary gland ultrasonography is commonly used to evaluate major salivary gland involvement in Sjogren’s disease (21). Although the relationship between lacrimal gland ultrasound(LGUS) characteristics and lacrimal gland (LG) histology remains unclear (22), a few studies have examined LGUS abnormalities, including fibrous gland appearance, heterogeneous texture, enlarged masses of cystic structures, and/or reticulated appearance (23–26). While studies have explored the application of ultrasound in evaluating conditions such as Sjogren’s disease, there is a notable absence of research on its utilization in the assessment of oGVHD. Furthermore, the correlation between lacrimal gland morphology and ocular surface clinical indicators in oGVHD remains unexplored. Compared with other evaluation tools mentioned in NIH CC 2014 or ICCGVHD criteria which focus on dry eye and ocular surface damage, lacrimal gland ultrasound may reduce the overlap with the diagnosis of dry eye disease and find some earlier changes in the lacrimal gland. Therefore, it is significant and possibly feasible to explore major lacrimal gland involvement by lacrimal gland ultrasonography in oGVHD. This study aims to investigate the effectiveness of lacrimal gland ultrasonography in assessing coGVHD following allogeneic hematopoietic stem cell transplantation (allo-HSCT) and determine the association between clinical ocular surface activity and ultrasonographic findings.

The procedures conducted for the study are consistent with the Helsinki Declaration and approved by the Ethics Committee of Peking University People’s Hospital. All subjects have been informed of the aim of the study, the principles of related examination methods and possible adverse consequences. Each individual signed an informed consent under total comprehension.

Fifty-seven patients (26 females, 31 males; age range from 18 to 65 years old; averaged 40.13 ± 11.074 years old) were included.

In this study, we evaluated the lacrimal gland involvement in oGVHD patients with ultrasonography. Our results suggest that lacrimal ultrasonography could be a non-invasive adjunctive tool for clinical assessment. Specifically, it proves valuable in evaluating the severity of dry eye, encompassing parameters such as tear secretion, as well as assessing the extent of corneal and conjunctival injury.

GVHD is the major complication following allo-HSCT and coGVHD emerges in 40-60% of patients. This condition is driven by complex interactions between the immune systems of the donor and the recipient, involving the recognition of host antigens by donor-derived CD4+ and CD8+ T cells (31). Dry eye disease is the hallmark of oGVHD and may be associated with inflammatory damage and fibrosis affecting the entire ocular surface system, including lacrimal and meibomian glands, cornea, conjunctiva, and eyelids (32, 33).

It has been established that the development of oGVHD is influenced by several factors linked to both donor and recipient characteristics, such as the human leukocyte antigen (HLA) mismatch or an unrelated donor, ABO mismatched and male recipients of female donors (34–36). However, in our study, aside from the duration after allo-HSCT, we did not observe significant differences in the aforementioned demographic and transplantation-related factors between oGVHD patients and non-oGVHD patients. This finding suggests that, in the context of chronic oGVHD, the duration after transplantation exerts a more pronounced impact on the deterioration of ocular surface conditions, including inflammatory damage and fibrosis while other factors mainly cause acute damage.

oGVHD most commonly involves changes to the ocular surface and can be characterized by keratoconjunctivitis sicca (or dry eye), as well as inflammatory damage to the conjunctiva, and punctate keratopathy (3, 10). It also involves inflammation of the lacrimal gland and eyelids, which is characterized by a decrease in tear secretion and tear film stability, as well as meibomian gland dropout (3, 37). In our study, parameters such as the OSDI, Schirmer test, TBUT, Tear meniscus height and corneal/conjunctival staining demonstrated significant differences between groups. This indicates that patients with oGVHD experience a reduction in tear secretion, impaired tear film stability and injury to the conjunctiva and cornea. However, the index of meibomian gland dropout and its grade did not show inter-group differences (p>0.05). This suggests that the impairment of the meibomian gland may occur in hematological patients already before HSCT, probably as the result of a multifactorial process caused by the concomitant therapies (i.e., chemo/radiotherapy) and/or the underlying disease itself with infiltration of the glands by tumor cells (37–39). As the majority of patients undergoing allo-HSCT have hematological malignancies, the occurrence of ocular toxicity induced by chemotherapeutic agents is noteworthy. Chemotherapy regimens can lead to a broad spectrum of ocular disorders such as dry eye, keratitis and lens disorders (40–42). Importantly, some of the mechanisms underlying such damage involve inflammatory reactions, similar to rejection. Thus, recognizing the inflammatory response’s role in ocular surface and lacrimal gland involvement underscores the impact of radiotherapy and chemotherapy, highlighting their significance in the context of ocular complications.

B-mode US can yield information on homogeneity, echogenicity, borders and parenchymal changes, such as intraparenchymal lymph nodes (30). Common parameters in lacrimal ultrasound are size, echogenicity, texture and parenchymal vascularization (27, 44). Although lacrimal ultrasound has been used in the assessment and even diagnosis of pSS, there is no related research applied in oGVHD. In our study, we apply B-mode and Doppler ultrasonography to evaluate the morphology, homogeneity and parenchymal vascularization. The confounding variables such as age and gender were matched between the groups. Most of the clinical indicators of the eyes were significantly different, but there was no significant difference in the results of lacrimal gland ultrasonography, which indicates that ocular rejection did not affect the morphology of the lacrimal gland under ultrasound. Here are some possible interpretations for the no difference between groups. Firstly, fibrosis and inflammation caused by stromal fibroblasts with T‐cell infiltration centers around the periductal area of the lacrimal gland are the main reasons for lacrimal gland dysfunction in oGVHD (43). However, the fibrosis is a relatively chronic process. Since studies that have examined lacrimal gland ultrasound abnormalities mentioned SjD develops over time and establishing a typical clinical picture typically takes years, the patients they enrolled turned out to have a long disease duration of years (>4 years 52.2%), which is significantly longer than the disease duration of patients enrolled in our study. Secondly, the epithelial‐mesenchymal transition of the host cells and the fibrosis progress may be triggered by chemoradiotherapy therapy before and after the HSCT (44). It suggests that the lacrimal gland changes in morphology could also be influenced by factors other than oGVHD such as chemoradiotherapy. Although there was no significant difference between GVHD and non-GVHD patients, which may suggest that ultrasound has no good effect on the diagnosis of GVHD, the morphology examined by ultrasound including long diameter and thick diameter showed a correlation with OSDI, tear meniscus height and corneal/conjunctival lissamine green staining, indicating that ultrasound can be meaningful for clinical evaluation by indicating the severity of dry eye, tear secretion and keratoconjunctival injury. However, ultrasound parameters didn’t show a correlation with TBUT, which suggests that it can’t reflect tear film stability. In addition to Doppler ultrasound, some studies have also explored the indicators of lacrimal gland under ultrasound 2-dimensional shear wave elastography (2D-SWE), a more effective imaging technique for evaluating tumor/nodule lesions and parenchymal fibrosis. It found a good correlation and diagnostic value (26). It suggests that we can use different ultrasound imaging techniques to evaluate the lacrimal gland.

Our study has several limitations. Firstly, the present study only showed results in a cross-sectional manner. Thus, the importance of LGUS on disease progression or change of LGUS was not evaluated. As a result, we cannot definitively ascertain whether lacrimal gland injury is attributable to radiotherapy, chemotherapy, or rejection. Secondly, the sample size was relatively small. Our analysis showed the homogeneity and parenchymal vascularization parameters were not significantly correlated with ocular surface and tear assessments. However, power analysis indicated that at a two-sided α level of 0.05 and with 80% power, the current analysis could only detect correlation coefficients with absolute values ≥0.36 due to the limited sample size (data not shown). Lastly, we only divided the grading of ultrasonic indicators including homogeneity and parenchymal vascularization into normal/abnormal, and did not form a detailed grading standard for accessing.

In conclusion, our study findings suggest a correlation between the morphology of lacrimal glands and clinical severity indices in oGVHD. Consequently, Doppler ultrasonography emerges as a potential auxiliary tool for early-stage clinical assessments of oGVHD, particularly in cases presenting with extra-glandular organ involvements and borderline diagnostic findings. This suggests that Doppler ultrasonography holds promise in enhancing diagnostic precision and timely intervention for individuals at the onset of oGVHD, contributing to improved patient outcomes and management strategies.