genotype–phenotype correlation in lattice corneal dystrophy: added diagnostic value of high-resolution OCT Authors: Katarzyna Walasz, Ewa Wróblewska-Czajka, Adam Wylęgała, Bartłomiej Bolek, Bogdan Dugiełło, Magdalena Kijonka, Anna Nowińska, Edward Wylęgała Published: 2026 ABSTRACT AbstractIntroduction: Lattice corneal dystrophy (LCD) is a progressive genetic disorder involving stromal amyloid accumulation. While molecular genetics is the diagnostic gold standard, its accessibility is often limited. This study evaluates high-resolution anterior segment optical coherence tomography (HR-OCT) for LCD characterization and surgical planning. e analysed 30 patients with confirmed TGFBI mutations (R124C, H626R) using slit-lamp biomicroscopy, in vivo confocal microscopy (IVCM) (Heidelberg Engineering GmbH, Heidelberg, Germany), and 3 μm resolution REVO HR-OCT (Optopol Technology, Zawiercie Poland). Results: IVCM visualized “felt-like” amyloid filaments at the cellular level. HR-OCT provided precise mapping of deposit depth, epithelial mapping and identified Bowman’s layer disruption, explaining clinical recurrent erosions. In post-keratoplasty cases, HR-OCT demonstrated high sensitivity in detecting early recurrence within donor tissue, often before changes became visible during slit-lamp examination. Conclusions: HR-OCT provides enhanced tool for LCD phenotypic characterization, serving as a “pre-operative roadmap” to differentiate candidates for phototherapeutic keratectomy versus keratoplasty. While IVCM offers cellular details, the wide-field imaging of HR-OCT is highly useful and significant for longitudinal monitoring and comprehensive management of LCD.Subject terms: Diseases, Medical research Introduction The differential diagnosis of corneal dystrophies is often challenging and was previously mainly limited to slit-lamp examinations until anterior segment optical coherence tomography (AS-OCT) was developed1. While molecular genetics remains the gold standard for definitive classification, access to such testing is frequently restricted by high costs, lack of insurance coverage, and the scarcity of specialized laboratories2. Consequently, AS-OCT has emerged as a crucial, widely accessible alternative3. Zeiss Visante OCT (Carl Zeiss Meditec) and the Slit-Lamp OCT (SL-OCT, Heidelberg Engineering) were the first commercially available AS-OCT devices. These systems utilized time domain OCT technology, which imposed limitations on resolution and A-scan/second rates4,5. The first anterior segment swept-source OCT, the Casia (SS-1000 OCT, Tomey) was introduced in 2008 and achieved significantly higher A-scan rate of 30,000 scans per second and an enhanced resolution of approximately 10 μm6. Crucially, this technology allows for the detailed visualization of all corneal layers, from the epithelium and Bowman’s layer to the stroma and endothelium. This capability is vital for pre-operative planning, as it enables the precise measurement of deposit depth. This data assists surgeons in choosing between minimally invasive procedures, such as phototherapeutic keratectomy7, and more extensive interventions like lamellar keratoplasty8. Understanding the genetics of dystrophy provides a critical foundation for comprehending the disease mechanism, paving the way for the innovation of precision medicines and gene-based interventions9,10. Lattice corneal dystrophy (LCD) is a bilateral and progressive autosomal dominant stromal disorder, primarily associated with mutations in the TGFBI gene11. It is characterized by the extracellular accumulation of amyloid deposits within the corneal stroma. These deposits manifest clinically as pathognomonic branching refractile lines and superficial fleck-like opacities12. Lesions typically originate in the central and superficial layers, subsequently spreading centrifugally and deeper into the corneal stroma over time. Patients frequently experience progressive visual impairment and painful recurrent corneal erosions, which often result from the mechanical disruption of Bowman’s layer by the enlarging amyloid aggregates13. Multimodal clinical imaging has become indispensable for precise diagnosis14. The integration of high-resolution optical coherence tomography (HR-OCT) and in vivo confocal microscopy (IVCM) enables detailed, non-invasive phenotypic characterization and precise localization of deposits. This study evaluates genetically confirmed lattice corneal dystrophy using a multimodal diagnostic approach, including confocal microscopy; however, HR-OCT emerged as the useful tool for detailed phenotypic characterization. Materials and methods Study population and patient selection From the original cohort of 41 LCD individuals characterized in our previous genetic study15 we selected a subgroup of 30 symptomatic patients for advanced morphological analysis. Informed consent was obtained from all patients before any study procedure or examinations were performed. The inclusion criteria were: (1) a molecularly confirmed mutation in the TGFBI gene associated with LCD, (2) presence of characteristic bilateral corneal opacities on slit-lamp examination, and (3) availability of high-quality multimodal imaging data (HR-OCT and IVCM). The study group consisted of 30 patients. Among them, 11 patients had a history of prior keratoplasty (penetrating or deep anterior lamellar); in these cases, imaging was focused on the graft or the recurrence of deposits at the graft-host interface. Exclusion criteria included other coexisting corneal pathologies (e.g., severe keratitis or significant scarring unrelated to LCD) that could interfere with imaging quality. The study was approved by the Ethics Committee of the Medical University of Silesia (SUM NN-6501–175/I/07) and conducted according to the Declaration of Helsinki. Clinical examination and genetic verification All subjects underwent a comprehensive ophthalmic examination, including Best-Corrected Visual Acuity (BCVA) measurement and slit-lamp biomicroscopy with digital photography (Topcon SL-D701). The clinical diagnosis of LCD was based on the presence of branching, linear stromal opacities. Genetic verification was performed using the protocol described previously15 which involved direct sequencing with the BigDye Terminator v3.1 (Applied Biosystems; Life Technologies, Foster City, CA) of the TGFBI gene to identify specific pathogenic variants (R124C and H626R). High-resolution optical coherence tomography Anterior segment imaging was performed using the REVO HR-OCT (Optopol Technology, Zawiercie, Poland). This Spectral Domain OCT (SD-OCT) utilizes a super-luminescent diode (870 nm) with an ultra-high axial resolution of 3 μm and a scanning speed of 130,000 A-scans/s. For each patient, we employed the Anterior line 9 mm scan to precisely measure the depth of amyloid deposits within the corneal layers, and the Anterior radial scan for epithelial thickness mapping. All scans were performed by a single experienced operator under consistent lighting conditions. The images were analysed for the presence of hyperreflective spindle-shaped or linear structures within the stroma. In vivo confocal microscopy To visualize the cornea at the cellular level, the HRT3 RCM (Heidelberg Engineering GmbH, Germany) was used. This confocal scanning laser microscope (670 nm diode laser) provides a 400 × 400 μm field of view with lateral and axial resolutions of 1 μm and 4 μm, respectively. The examination procedure was standardized as follows: Topical anaesthesia was induced with 0.5% proparacaine hydrochloride. A sterile disposable PMMA cap (TomoCap) was filled with a synthetic tear gel (carbomer) as a coupling medium and placed on the objective lens. The central cornea was scanned from the epithelium down to the endothelium. Specific attention was paid to the subepithelial plexus and mid-stroma to identify the pathognomonic hyperreflective, branching “felt-like” filaments. A minimum of 50 images per eye were captured to ensure a representative assessment of the lattice structures. Lattice corneal dystrophy grading Based on the findings from slit-lamp biomicroscopy, IVCM and HR-OCT, patients were categorized into three clinical stages (Table 1): Table 1Multimodal classification of LCD severity stages.StageSLIT-LAMPIVCMHR-OCTEarlydistinct subepithelial ovoid dots, transparent branching refractile filamentsinitial branching filaments in the subepithelial or superficial stromasparse hyperreflective deposits limited to superficial stroma; Bowman’s layer remains intactModeratethin, but clearly visible branching refractile lines, whitish, subepithelial dotsprogression of branching filaments localized in the anterior to mid-stroma with early disruption of Bowman layerwell-defined hyperreflective structures in the anterior and mid-stroma; focal disruptions of Bowman’s layerAdvanceddense lines spreading centrifugally and deeper into the stroma; Bowman layer disruption, epithelial layer thickness variations, superficial erosions and significant loss of corneal transparencyextensive network of filaments with blurred margins throughout the stroma; presence of areas consistent with hyperreflective, irregular material suggesting fibrosisdense, linear deposits throughout the entire stroma; significant Bowman’s disruption, epithelial layer irregularities, and diffuse stromal hyperreflectivity suggesting fibrosis Multimodal classification of LCD severity stages. Results Slit lamp biomicroscopy Examination in slit lamp biomicroscopy revealed central, superficial fleck-like opacities. These lesions progressively evolved into thin, branching refractile lines and whitish, ovoid subepithelial dots. The pathology followed a distinct distribution pattern, originating centrally and superficially before spreading centrifugally and deeper into the stroma16. In vivo confocal microscopy IVCM revealed distinct structural alterations in LCD, characterized by deposits arranged in a branching, lattice-like pattern. These deposits were predominantly located in the stroma. The characteristic “felt-like” appearance of these branching, high-contrast filaments correspond to the amyloid lattice lines observed in clinical biomicroscopy, providing a cellular-level visualization of the disease’s stromal progression17. High-resolution anterior OCT High-resolution AS-OCT imaging provided a detailed cross-sectional analysis of the corneal structural alterations in both unoperated and post-keratoplasty patients, offering significant added value when correlated with slit-lamp biomicroscopy and IVCM. In cases of advanced LCD, where slit-lamp examination (Fig. 1c) revealed dense, confluent opacities and a global loss of transparency, HR-OCT visualized a dense distribution of multiple, high-intensity hyperreflective linear deposits throughout the entire stromal thickness (Fig. 3a, red arrows). These amyloid aggregates were associated with a significant loss of integrity in Bowman’s layer (Fig. 3a yellow arrow) and subsequent epithelial irregularities (Fig. 3a, green arrow)18. Furthermore, extensive subepithelial and anterior stromal fibrosis was observed, contributing to the overall loss of corneal transparency. These findings responded to IVCM, which visualized an extensive network of filaments with blurred margins and areas with high reflectivity suggestive of fibrosis (Fig. 2a-d). Representative slit-lamp biomicroscopy images illustrating the spectrum of corneal phenotypic features observed in the lattice corneal dystrophy. (a–c) Central, superficial, fleck-like opacities progressively evolving into thin, branching refractile lines and whitish, ovoid subepithelial dots (red arrow). (d) Cornea of a patient after keratoplasty. The central corneal graft (red star) exhibits scarring, while characteristic branching lattice lines (red arrow) persist in the peripheral host cornea (blue star), outside the graft-host junction. (e) Advanced lattice corneal dystrophy. (f) Recurrence of lattice corneal dystrophy manifesting as anterior stromal opacification within the donor tissue. High-resolution anterior segment optical coherence tomography imaging of morphological changes in lattice corneal dystrophy. (a-b) Unoperated patients: (a) Advanced LCD: Numerous hyperreflective linear deposits in the corneal stroma (red arrows); significant disruption of Bowman’s layer (yellow arrow); epithelial irregularity (green arrow); subepithelial and anterior stromal fibrosis (blue star). (b) Moderate LCD: Characteristic amyloid deposits within the stroma (red arrows); Bowman’s layer disruption (yellow arrow). (c-d) Post-keratoplasty patients: (c) Distinct amyloid deposits in the recipient cornea (blue arrow); graft-host junction (green star); recurrence of dystrophy within the donor graft (red arrows); Bowman’s layer disruption associated with subepithelial scarring (yellow arrow). (d) Findings consistent with panel C; recurrence of dystrophic changes within the donor tissue (red arrows); amyloid deposits in the peripheral host cornea (blue arrow) beyond the graft-host junction (green star); Bowman’s layer disruption associated with subepithelial scarring (yellow arrow). Representative in vivo confocal microscopy images showing the cellular morphology of the corneal stroma in lattice corneal dystrophy. (a–d) Distinct structural alterations in LCD, characterized by hyperreflective, blurred margins deposits arranged in a branching, lattice-like pattern (red arrows). Artifact (green dot). Disorganised areas suggestive of fibrosis (blue star). Hyperreflective stromal architecture suggestive of scarring (yellow triangle). In patients with moderate LCD, biomicroscopy findings typically included fine refractile filamentary lines and discrete ovoid subepithelial opacities (Fig. 1b). HR-OCT localized these deposits as well-defined hyperreflective structures primarily within the anterior and mid-stroma, with focal disruptions of Bowman’s layer even in areas that appeared clinically stable (Fig. 3b, red and yellow arrows). In the subgroup of patients after keratoplasty, HR-OCT allowed for precise visualization of the graft-host interface19. While slit-lamp biomicroscopy often showed non-specific graft scarring or persistent peripheral lattice lines in the host cornea (Fig. 1d), HR-OCT enabled detection of subclinical recurrence within the donor tissue (Fig. 3c, d). In these cases, the imaging confirmed the persistence of pathognomonic lattice deposits within the peripheral host rim, extending beyond the graft junction. More importantly, HR-OCT identified the recurrence of the dystrophy within the donor tissue, manifested as new hyperreflective deposits and subepithelial scarring (Fig. 3c, d). Epithelial thickness mapping Epithelial thickness mapping disclosed critical insights into the ocular surface instability characteristic of LCD. In the acute phase of the disease, HR-OCT visualized areas of significant epithelial compromise, including recurrent erosions (Fig. 4a). These regions were characterized by focal epithelial absence or extreme thinning (down to 16 μm), coinciding with marked irregularities in both Bowman’s layer and the anterior stroma. Furthermore, the tear film profile in these zones exhibited significant instability (Fig. 4a, blue arrow). Epithelial remodeling and ocular surface stability in advanced lattice corneal dystrophy. (a) Acute phase: erosion. Regions of extreme focal epithelial thinning (down to 16 μm) coincide with marked irregularities in both Bowman’s layer and the anterior stroma. The blue arrow indicates significant tear film instability in the affected zone. (b) Follow-up imaging of the same patient. Compensatory epithelial hypertrophy (124 μm) is observed at the previous erosion site, effectively stabilizing the corneal surface, while underlying stromal thinning (326 μm) remains evident. (c–d) Automated epithelial thickness maps corresponding to the acute (c) and remodeling (d) phases. Focal thinning and compensatory hypertrophy are reliably identified in corresponding corneal locations across both analysis methods. Follow-up imaging of the same patient after clinical resolution revealed a dynamic process of epithelial remodeling (Fig. 4b). This stage was characterized by compensatory epithelial hypertrophy (124 μm) at the previous erosion site, which effectively “filled” the structural gap and stabilized the corneal surface. Notably, while the surface regularity improved, underlying stromal thinning (326 μm) was observed in these areas. It should be noted that while manual measurements on cross-sectional scans (Fig. 4a–b) exhibited minor numerical discrepancies compared to the values generated by automated epithelial mapping (Fig. 4c–d), a consistent trend in morphological alterations was observed. Specifically, focal thinning and compensatory hypertrophy were reliably identified in corresponding corneal locations across both analysis methods, ensuring the clinical validity of the observed epithelial remodeling. Depth distribution of stromal deposits To illustrate the clinical utility of HR-OCT in surgical planning, representative depth measurements of the stromal deposits were performed (Fig. 5a-b). In cases clinically graded as moderate (Fig. 5a), hyperreflective lesions were observed at depths of approximately 118 μm, typically remaining within the anterior stroma. In contrast, in advanced stages (Fig. 5b), deposits were identified reaching as deep as 466 μm. These parameters demonstrate the capability of HR-OCT to objectively distinguish between superficial and deep stromal involvement, thereby facilitating the selection of the appropriate surgical approach. Depth measurements of stromal deposits for surgical planning. (a) Moderate LCD: hyperreflective amyloid lesions localized within the anterior stroma at a measured depth of approximately 118 μm. (b) Advanced LCD: deposits are identified extending significantly deeper into the corneal architecture, reaching a depth of 466 μm. Longitudinal monitoring of post-keratoplasty changes In the post-keratoplasty group, longitudinal observations demonstrated a progression in both the quantity and depth of recurrent amyloid lesions over a 1-year period. A clear trend was observed where recurrent deposits were no longer restricted to the graft-host interface. HR-OCT revealed that recurrent hyperreflective spindle-shaped lesions were identified both within the central donor stroma (Fig. 6a-c, red square) and at the donor-recipient junction (Fig. 6a-c, red arrows). The deposits at the interface were clearly visualized and exhibited high optical reflectivity. While the exact pathophysiology of recurrence—whether originating from the graft-host interface or arising de novo within the donor tissue—remains unclear, HR-OCT facilitated a detailed record of the evolving corneal architecture. Despite the increasing stromal involvement, the overlying epithelium remained relatively stable during the follow-up period. Further monitoring is required to assess the long-term impact on graft clarity. Longitudinal (1 year follow-up) HR-OCT monitoring of recurrent amyloid lesions in a post-keratoplasty patient. (a–c) The progression of LCD in both quantity and depth. To ensure morphological consistency across follow-up visits, each scan was performed along the same horizontal axis (180° through the central cornea). Recurrent hyperreflective spindle-shaped lesions (red arrows) are identified within the central donor stroma (red square) and at the donor-recipient junction (green square). Discussion Corneal diseases affect millions worldwide, requiring precise diagnostic tools for effective management20. Traditionally, IVCM has been the gold standard for in vivo cornea visualization, offering high-resolution images crucial for evaluating corneal innervation and other microscopic structures. However, the recent advent of HR-OCT introduces a powerful alternative, promising detailed imaging of corneal and anterior segment structures21,22. HR-OCT is a powerful diagnostic tool that aids in distinguishing ocular surface disorders by offering detailed cross-sectional images that reveal specific, anatomic changes. Visualizing individual corneal layers and precisely locating lesions with HR-OCT improves the accuracy of lesion assessment by enabling a more precise assignment of hyperreflectivity to specific structures. This study is the first to compare images of lattice stromal dystrophy captured through slit-lamp microscopy, confocal microscopy, and high-resolution OCT, offering a valuable opportunity to improve diagnostic accuracy in these disorders. Our primary emphasis was on assessing HR-OCT images and their significance in guiding diagnosis. Future research could focus on implementing automated quantitative mapping of amyloid load in the posterior stroma and investigating the impact of deep stromal deposits on corneal biomechanical properties and visual quality (e.g., corneal densitometry or higher-order aberrations). There is a growing need for devices that offer non-invasive, high-resolution imaging capable of visualizing both corneal and anterior segment structures comprehensively23. Furthermore, while IVCM offers a restricted field of view (400 × 400 μm), the 9 mm scan length of the HR-OCT provides a comprehensive overview of the distribution pattern—from the central cornea to the periphery—making it more effective for assessing the “centrifugal” progression of lattice dystrophy. This difference is particularly critical when evaluating the integrity of Bowman’s layer. Because the disruption of this layer in the moderate stage of LCD is often focal or segmental, these pathological changes may be missed by IVCM if they occur outside the immediate scanning area. In such cases, the wide-field cross-sectional imaging of HR-OCT allows for more reliable detection and mapping of these irregularities across the entire cornea. Other research24 highlights the importance of utilizing multiple imaging modalities, including slit-lamp examination, HR-OCT, in vivo confocal microscopy, and UBM, to achieve the most precise diagnosis of anterior segment disorders in the eye. Traditionally, the assessment of deposit depth for surgical planning in LCD can be achieved through slit-lamp biomicroscopy25 or Scheimpflug imaging26,27. Slit-lamp examination allows for a subjective estimation of depth; however, it lacks the numerical precision required for micron-level planning of phototherapeutic keratectomy (PTK). While modern Scheimpflug systems like Pentacam Cornea OCT (Oculus, Germany) have evolved to provide high-resolution axial imaging and automated densitometry maps, further research is warranted to directly compare their diagnostic efficacy with HR-OCT in the context of lattice corneal dystrophy. Future studies involving larger, prospective patient cohorts could establish how these two complementary technologies might be integrated to optimize surgical planning and long-term monitoring. Such a comparison would be particularly valuable for validating standardized ‘cut-off’ values for phototherapeutic keratectomy versus keratoplasty28. Regarding surgical strategy, we do not advocate for the ablation of all visible amyloid deposits if they extend into the deep stroma. It could lead to excessive corneal thinning and significant hyperopic shifts29, compromising the eye’s refractive stability. By analyzing the reflectivity profile, surgeons can target the most optically dense anterior deposits that cause the greatest light scattering, while preserving a safe residual stromal bed30,31. A major strength of this study is the longitudinal analysis of 11 patients who underwent keratoplasty. Our findings demonstrate that HR-OCT is more precise to slit-lamp biomicroscopy in the early detection of lattice dystrophy recurrence. We observed that hyperreflective spindle-shaped deposits often emerge at the graft-host junction or in the subepithelial space of the donor tissue before they become clinically visible as opacities. This early visualization enables clinicians to monitor the recurrence rate more closely and consider early intervention with PTK to maintain graft clarity32,33. Our imaging results revealed a clear link between the mechanical disruption of Bowman’s layer and epithelial irregularities. These morphological changes provide a physical explanation for the recurrent corneal erosions and chronic pain frequently reported by LCD patients. The ability of HR-OCT to visualize amyloid aggregates “pushing” through Bowman’s layer into the epithelium allows clinicians to better explain the pathophysiology of pain to patients and tailor management strategies, such as the use of therapeutic bandage lenses or targeted laser treatment34. Conclusions This study demonstrates that HR-OCT is a useful tool for the phenotypic characterization of lattice corneal dystrophy, offering higher precision than traditional slit-lamp examination. Its ability to accurately map the depth of amyloid deposits makes it a highly valuable pre-operative roadmap for choosing between PTK and keratoplasty. Furthermore, HR-OCT enables the early detection of recurrence in post-keratoplasty patients before changes become clinically visible. By visualizing the disruption of Bowman’s layer, it also provides a clear morphological explanation for recurrent erosions and pain. Ultimately, while IVCM offers cellular details, the wide-field imaging of HR-OCT is indispensable for the comprehensive management and longitudinal monitoring of LCD patients. Author contributions Katarzyna Walasz and Bogdan Dugiełło were involved in the writing of the manuscript, study design, and acquisition of data using the OCT devices. Bartłomiej Bolek and Magdalena Kijonka contributed to the patient recruitment analysis and interpretation of the data. Anna Nowińska, Adam Wylęgała and Ewa Wróblewska-Czajka were responsible for drafting the manuscript and revising it critically for intellectual content. Edward Wylęgała supervised the entire study, wrote the result sections contributed to data analysis, and provided final approval of the version to be published. Funding This research was supported by statutory funding from the Medical University of Silesia (BNW-1-130/K/3/K). Data availability The datasets generated or analysed during the current study are not publicly available due to patient privacy and GDPR regulations, but are available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets generated or analysed during the current study are not publicly available due to patient privacy and GDPR regulations, but are available from the corresponding author on reasonable request.