Retinal Structure and Microcirculation Alterations in Patients With Papillary Thyroid Carcinoma: An Optical Coherence Tomography and Angiography Study

Abstract

Purpose: This study aimed to compare the retinal thickness, vessel density, and perfusion density in patients with papillary thyroid carcinoma (PTC) and healthy controls using optical coherence tomography (OCT) and OCT angiography. And investigating the influence of serological parameters and carcinoma signs with retinal structural and microcirculatory in PTC patients.

Methods: This cross-sectional study included 174 PTC patients (345 eyes) and 179 healthy subjects (358 eyes). Serological parameters (complete blood count, thyroid function, and lymphocyte subsets) and retinal thickness and flow parameters were compared between the two groups, and the correlations of retinal parameters with serological parameters in the PTC group. We investigated the effect of carcinoma signs (tumor size, focus location, lymphatic metastasis, surrounding tissue invasion, and BRAF mutations) on retinal parameters.

Results: PTC patients exhibited significantly thinner retinal thickness and reduced vessel density and perfusion density in the macular superficial vascular plexus compared with healthy controls. Platelet-to-lymphocyte ratios and thyroglobulin were inversely correlated with retinal parameters, and CD4⁺ T cells were positively correlated. Aggressive carcinoma signs thicken the retinal thickness more than nonaggressive cases.

Conclusions: The findings indicate a trend toward macular thinning and retinal microcirculatory dysfunction in the PTC patients, and these changes may be related to chronic inflammation and immune dysregulation secondary to cancer progression. OCT and OCT angiography show potential as noninvasive tools for detecting subclinical retinal abnormalities in PTC patients, and retinal alterations may serve as a surrogate marker for systemic inflammation. However, further studies are needed to address confounders and establish causality.

Over the past decades, the incidence of papillary thyroid carcinoma (PTC), which constitutes 90% of thyroid cancer cases,¹ has increased dramatically, reaching 25.8 cases per 1,000,000 among Chinese women.^2,3 Although PTC generally has a favorable prognosis, the proliferation and metastasis of cancer cells can lead to thyroid dysfunction as well as localized inflammatory and immune responses.⁴ Thyroid inflammatory triggers and persistently sustains autoimmune processes in the eye. Inflammatory factors damage retinal endothelial cells, leading to an imbalance in retinal vascular homeostasis.⁵ Significant decreases in retinal thickness and vascular perfusion have been observed in thyroid-associated ophthalmopathy and Graves' disease.^6,7 

Systemic inflammatory disease is increasingly recognized as significant risk factor for retinal maculopathy.^8,9 Photoreceptor atrophy, necrosis of the pigment epithelium in the macular area, and ischemic hypoxia are associated with an inflammatory response secondary to systemic disease that persists and damages the retina progressively.¹⁰ Studies of microvascular pathogenicity but without lesions have received attention. For example, retinal thickness and blood flow changes can be observed in the retinas of diabetic patients with undiagnosed retinopathy.¹¹ The impact of cancer on the incidence of AMD has also been widely developed.^12–14 Research indicates that the risk for macular disease is 1.38 to 1.92 times higher in patients with thyroid cancer than in individuals without cancer.¹⁴ However, detailed analyses of the retinal thickness and blood flow in thyroid cancer are lacking. Further investigation is needed to determine whether thyroid cancer induces systemic inflammation and alters blood flow, and whether these changes are associated with specific retinal alterations. 

Optical coherence tomography (OCT) and OCT angiography (OCTA) are noninvasive methods used for examining the retina in vivo. These technologies, enhanced by advanced computing, allow for the assessment of both static photoreceptor tissue and dynamic blood flow, enabling layered and segmented analysis of the fundus anatomy and perfusion. This approach helps clinicians in early diagnosis, disease monitoring, and prognostic evaluation. Accordingly, this study aimed to use OCT and OCTA to investigate quantitatively the retinal structural and microcirculatory characteristics in PTC patients and to investigate the relationships of serological inflammatory markers and carcinoma signs, respectively, with retinal changes. 

Methods

Study Cohort

This cross-sectional observational study was performed at the Department of Thyroid Surgery, Ophthalmology, and Health Examination of the First Hospital of Shanxi Medical University in Shanxi, China, between July 2023 and December 2023. Written informed consent was obtained from all participants. The study was approved by the Ethics Committee of First Hospital of Shanxi Medical University, and the methodology adheres to the tenets of the Declaration of Helsinki for research involving human subjects. 

We excluded patients who did not meet the criteria for this study (Fig. 1). The study included 174 patients with PTC (345 eyes), comprising 127 women (271 eyes) and 47 men (74 eyes). The healthy control (HC) group consisted of 179 individuals (358 eyes), with 125 women (250 eyes) and 54 men (108 eyes). 

Figure 1.

 

The selection of PTC patient data for analysis.

The inclusion criteria for patients with PTC were as follows: (1) a diagnosis of PTC confirmed via thyroid puncture biopsy, with no prior treatment history, including hormonal treatments; and (2) best-corrected visual acuity (BCVA) of >0.6 and IOP within the normal range (10–21 mm Hg). 

The inclusion criteria for the HC group were as follows: (1) no history of thyroid or ocular diseases; (2) no systemic or local use of antibiotics, immunosuppressive agents, or hormones; (2) no prior history of hypertension or diabetes mellitus were included, with blood pressure and fasting blood glucose levels confirmed to be within the normal physiological range (systolic blood pressure of <120 mm Hg and diastolic blood pressure of <80 mm Hg; fasting blood glucose, 3.9–6.1 mmol/L); (3) no systemic or local use of antibiotics, immunosuppressive agents, or hormones; (4) no significant systemic or local inflammation or immune diseases; and (5) BCVA of >0.6 and IOP within the normal range (10–21 mm Hg). 

The exclusion criteria for all the participants were as follows: (1) inability to undergo examination owing to poor vision, nystagmus, or refractive interstitial opacity; (2) optic nerve diseases or vitreoretinal interface disorders; (3) history of ocular trauma or surgery; (4) history of uveitis; (5) use of medications affecting retinal microcirculation; and (6) poor imaging quality, defined as an OCT and OCTA image scan quality index of <6. 

General Inspection Items

The common examination items were as follows. 

• (1) Personal information, including gender, age, drinking and smoking, and history of diabetes or hypertension.
• (2) Examination using a slit-lamp microscope to exclude cataract surgery, ocular trauma, uveitis, and other underlying ocular diseases. BCVA measurement (based on an international standard logarithmic visual acuity chart [converted to logMAR]), IOP measurement, spectral domain OCT imaging, and OCTA imaging were performed.
• (3) Blood samples were collected to measure thyroid function, including free triiodothyronine, free thyroxine, thyroid-stimulating hormone, thyroglobulin (Tg), anti-Tg antibody (TgAb), and antithyroid peroxidase autoantibody (TPOAb), using an automatic biochemical analyzer.
• (4) Blood samples were collected to test complete blood count, including leukocyte, platelets, neutrophils, monocytes, and lymphocytes, using an automatic flow cytometry blood cell counter. The following combined indices were also evaluated: neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR).
• (5) Blood samples from PTC patients were collected to determine lymphocyte subsets, including lymphocytes, T cell, CD4⁺ T, CD8⁺ T, B cell, natural killer (NK) cell, and CD4⁺ T/CD8⁺ T, using an automatic flow cytometer. Additionally, carcinoembryonic antigen concentration were measured.
• (6) Tumor tissues were excised during thyroid surgery to assess tumor size and focus location. Postoperative biopsies were conducted to determine tumor type, lymphatic metastasis, surrounding tissue invasion (including fibrous capsule, fat, intravascular cancer emboli, nerves, and muscle), and BRAF gene mutation status.

The OCT and OCTA Inspection Items

All participants underwent OCT and OCTA imaging using the Cirrus HD-OCT 5000 Review Software V.10 (Carl Zeiss Meditec Inc., Jena, Germany). The macula was divided into three rings and nine zones based on the Early Treatment of Diabetic Retinopathy Study (ETDRS). The rings encompassed 0 to 1 mm, 1 to 3 mm (internal ring), 3 to 6 mm (outer ring), and 0 to 6 mm, which correspond with the macular fovea, parafovea, perifovea, and whole en face, respectively. The macula was divided into four quadrants as superior, inferior, nasal, and temporal (Fig. 2). The thickness of the macula (the inner limiting membrane [ILM] to the RPE) and the ganglion cell layer (GCL) was scanned using the Macular Cube 512 × 128 program and calculated using the OU analysis mode. 

Figure 2.

 

OCT and OCTA image of the retina and division method. (a) The measurement of macula thickness (ILM-RPE) and blood flow using the Early Treatment Diabetic Retinopathy Study grid. (b) The measurement of GCL thickness partition diagram. (c) The range between the dotted lines indicates a superficial retinal vessel. IR, inner ring; OR, outer ring.

OCTA of the macular area was performed using the angiography 6 mm × 6 mm program. The vessel density (VD) and perfusion density (PD) were calculated using the built-in angiography analysis mode (FORUM platform, Carl Zeiss Meditec Inc.) using the optical microangiography algorithm, which isolates dynamic blood flow signals by detecting temporal variations in OCT signal caused by moving red blood cells. Nonperfusion artifacts, including speckle noise, were minimized by excluding pixel clusters of <6 pixels, and adaptive thresholding was used to differentiate perfused vessels from background noise, following protocols validated by the manufacturer. The resulting images were transformed into binary maps, with pixels representing flow signals assigned a value of 1 (vessel) and others 0 (nonvessel), facilitating quantitative evaluation. The VD was calculated as the proportion of the total scanned area occupied by binarized vessels within each ETDRS subfield, representing the structural density of the retinal vasculature. The PD was determined as the ratio of pixels with flow signals exceeding the noise threshold to the total pixels in each subfield, reflecting functional perfusion efficiency and hemodynamic activity. The VD primarily assesses vascular morphology, such as branching patterns and coverage, and PD evaluates hemodynamic performance, including blood delivery efficiency. This dual-parameter approach improves diagnostic precision: decreased VD may indicate vascular atrophy or occlusion, reduced PD suggests perfusion impairment, and elevated PD could reflect compensatory hyperperfusion or neovascular responses. 

Because the software version of the instrument used did not support analysis of the VD and PD of deep layer retinal vessels, only VD and PD data for the macular superficial capillary plexus (from the ILM to the inner plexiform layer) were obtained. All the procedures were performed by the same skilled technician (Z.-Y.Z.). Retinal specialists performed manual adjustments to correct errors in foveal positioning and retinal stratification to ensure the accuracy of all image segmentation (T.-Y.L.). 

Statistical Analysis

All statistical analyses were conducted using IBM SPSS Statistics 25 (IBM Corporation, Armonk, NY, USA). Baseline characteristics are presented as means ± SDs (for continuous variables) or frequencies and percentages (for categorical variables). Continuous variables were analyzed using the independent samples t test, and categorical variables were compared using the χ² test. Spearman's rank correlation was used to assess the relationship between continuous variables. The generalized estimating equation model was used to correct the correlation of retinal structural and microcirculatory parameters in both eyes. A significance level of α = 0.05 was set, with a P value of <0.05 indicating statistical significance. 

To address multiple comparisons in the analysis of retinal parameters, Bonferroni correction was applied to control the family-wise error rate. Stratified Bonferroni correction was implemented as follows. 

• 1. Rationale for multiple comparisons correction.

Anatomical independence does not equate to statistical independence. Although ETDRS sectors are distinct anatomically, retinal regions may be influenced by shared systemic factors (e.g., inflammation, hemodynamics), resulting in statistical interdependence. For instance, without correction, the cumulative false-positive risk for nine comparisons of macular (ILM-RPE) thickness is approximately 37% (1 − (1 − 0.05)⁹ ≈ 0.37). 

• 2. The significance thresholds were adjusted according to the number of comparisons within each parameter category as follows.

ILM-RPE Thickness (9 sectors): α^a = 0.05/9 = 0.0056; GCL thickness (6 sectors): α^b = 0.05/6 = 0.0083; whole en face VD/PD (4 sectors): α^c = 0.05/4 = 0.0125; quadrant VD/PD (8 sectors): α^d = 0.05/8 = 0.00625; with the P value of less than the adjusted α maintaining statistical significance. 

Results

Characteristics of the Study Participants

This study involved 353 participants (703 eyes), consisting of 174 patients (345 eyes) with PTC and 179 HC (358 eyes), matched by gender and age. The demographic characteristics and baseline clinical features of the PTC and HC groups are shown in Table 1. No significant differences were found between the two groups in age, sex, IOP, and drinking history (P > 0.05). Significant differences were observed in BCVA (P < 0.001) and smoking history (P = 0.018). The proportion of cigarette smokers was greater in the PTC group (10.34%) than in the HC group (3.91%), and the mean BCVA was slightly lower in the PTC group (0.20 vs. 0.15). 

Table 1.

 

Demographic and Clinical Characteristics Between PTC and HC Groups (Mean ± SD)

 

The hematological parameters of the participants are shown in Table 2. Serum Tg, TgAb, and TPOAb levels, as well as NLR and PLR, were significantly higher in the PTC group than in the HC group. Additionally, the absolute monocyte count in the PTC group were significantly lower than those in the HC group (all P < 0.05). 

Table 2.

 

Hematological Parameters Between PTC and HC Groups (Mean ± SD)

 

Comparison of Retinal Structural and Flow Parameters Between the PTC and HC Groups

The PTC group exhibited a significantly thinner retinal thickness than the HC group. Statistically significant differences in macular thickness were observed in the inner ring (inferior and nasal) and outer ring (inferior, nasal, and temporal) areas (all P < 0.05). However, following Bonferroni correction for multiple comparisons (adjusted α^a = 0.0056), only the nasal and temporal regions of the outer ring (all P < 0.001) maintained statistical significance (all corrected P < 0.0056) (Table 3, Fig. 3a). The GCL thickness was significantly thinner in the PTC group in the temporal sectors, respectively, the superotemporal and inferotemporal regions (both P < 0.001). These differences remained statistically significant after Bonferroni correction for multiple comparisons (adjusted α^b = 0.0083, all corrected P < 0.0083) (Table 3, Fig. 3b). 

Table 3.

 

OCT-Based Comparison of Retinal Thickness Parameters Between PTC and HC Groups (Mean ± SD)

 

Figure 3.

 

Columnar plots of retinal thickness and flow parameters between PTC and HC groups. IR, inner ring; OR, outer ring. *P < 0.05; after Bonferroni correction. ^aP < 0.0056, ^bP < 0.0083, ^cP < 0.0125, and ^dP < 0.00625.

The alterations in retinal blood flow parameters exhibited divergent trends. In the whole analysis, statistically significant differences in the VD and PD were observed in the fovea, parafovea, perifovea, and whole en face (all P < 0.05), except for parafoveal PD (P = 0.051). Notably, foveal PDs were significantly increased, whereas the other parameters showed a decrease. After implementing Bonferroni correction for multiple comparisons (adjusted α^c = 0.0125), significant differences were retained in fovea and parafovea VD, as well as perifovea and whole en face PD (all corrected P < 0.0125) (Table 4, Fig. 3c). In a quadrant analysis, significant decreases were observed in the VD of the temporal outer ring, and PD of the inferior inner ring and all outer ring quadrants (all P < 0.05). VD in the superior and inferior quadrants of the outer ring were significantly higher than the HC group (both P < 0.05). After Bonferroni correction (adjusted α^d = 0.00625), significant changes persisted in VD of the superior and temporal quadrants, and PD of the inferior, nasal, and temporal quadrants of the outer ring (all corrected P < 0.00625) (Table 4, Fig. 3d). The between-group differences in retinal parameters were most pronounced in the temporal and outer ring regions, especially after Bonferroni correction (Fig. 4). 

Table 4.

 

OCTA-Based Comparison of Macular Flow Parameters Between PTC and HC Groups (Mean ± SD)

 

Figure 4.

 

Plot of regional differences between PTC and HC groups. Adjusted α^a = 0.0056, α^b = 0.0083, α^c = 0.0125 and α^d = 0.00625.

Effect of Blood Markers on Retinal Thickness and Flow Parameters in the PTC Group

The correlations between blood levels of inflammatory markers and retinal thickness and flow parameters are presented in Table 5 and Figure 5. Platelets and PLR showed negative correlations with retinal thickness, and NK cells and carcinoembryonic antigen concentration showed negative correlations with retinal flow (all P < 0.05). Conversely, Tg, free triiodothyronine, lymphocytes, monocytes, and CD8⁺ T% exhibited positively correlated with multiple rings and quadrants of retinal thickness. Similarly, CD4⁺ T cell and CD4⁺ T/CD8⁺ T cell ratios were positively correlated with specific rings and quadrants of PD or VD. 

Table 5.

 

Spearman's Rank Correlation Coefficient of Inflammatory Markers With Retinal Structural and Flow Parameters

 

Table 6.

 

Effect of Carcinoma Signs on Retinal Thickness and Flow Parameters

 

Figure 5.

 

Heatmap of correlations between blood markers and retinal thickness and flow parameters. CEA, carcinoembryonic antigen; FT₃, free triiodothyronine; FT₄, free thyroxine; IR, inner ring; LYMPH, lymphocytes by lymphocyte subset testing; OR, outer ring; Tg, thyroglobulin; TgAb, antithyroglobulin antibody; TSH, thyroid-stimulating hormone. The y axis represents, from top to bottom: thyroid function parameters (1–6), complete blood count (7–14), and lymphocyte subpopulation parameters (17–29).

Effect of Carcinoma Signs on Retinal Thickness and Flow Parameters in the PTC Group

The carcinoma signs of 168 PTC patients (334 eyes) were documented in detail. Among them, 52 patients had tumor size of ≥1 cm, 45 had a bilateral focus, 95 had lymphatic metastasis, 114 had surrounding tissue invasion, and 86 had BRAF mutations. The analysis of ocular parameters across different carcinoma signs subgroups revealed several significant changes (Table 6). In patients with tumor size of ≥1.0 cm, the nasal outer ring VD and the outer ring (superior and nasal) PD were significantly increased compared with those with smaller tumors (all P < 0.05). Lymphatic metastasis was associated with increased retinal thickness in the inner ring (nasal and temporal, all P < 0.05). Tumors with surrounding tissue invasion exhibited greater thickness of the inner ring (nasal and temporal) and the nasal outer ring (all P < 0.05). BRAF mutation carriers increased thickness in the inferior outer ring (P = 0.013). In contrast with these changes, bilateral focus had reduced retinal thickness inferior outer ring (P = 0.018). However, all results lost significance after the Bonferroni correction. 

Discussion

The recognition that macular structural and microcirculatory alterations occur during the subclinical phase of retinal disease has gained increasing attention. Consistent with emerging evidence, we believe that systemic diseases—particularly those associated with chronic inflammation or immune dysregulation—can exert profound effects on retinal structure and microcirculation. OCT and OCTA, although developed later in ophthalmology, have become essential, noninvasive, convenient, safe, and efficient strategies for assessing retinal structure and microcirculation, especially in linking retinal and systemic diseases.¹⁵ Lin et al.¹⁴ have shown that the PTC population is associated with a higher risk of macular degeneration than the healthy population. In this study, we quantitatively analyzed the retinal structure and microcirculation in PTC patients using OCT and OCTA. The results demonstrated that the macula was significantly thinner in PTC patients compared with healthy individuals, accompanied by impaired macular VD and perfusion. These changes were localized predominantly to the temporal and outer ring (perifoveal) regions, which may serve as early biomarkers for identifying underlying macular pathology. 

The systemic chronic disease known to have the most widespread effect on the retina is diabetes mellitus, followed by hypertension. Individuals who have been diagnosed with diabetic or hypertensive retinopathy may present with macular atrophy or edema, retinal microaneurysms, or vitreous hemorrhage. The retinal characteristics of individuals who have abnormal blood glucose or blood pressure alone, but have not yet developed retinopathy, have also been studied extensively. Studies have shown that the outer nuclear layer and myoid zone of the retina were thinned in patients with prediabetes,¹¹ and hyperglycemic patients without retinopathy had attenuated PD and VD in both the superficial and deep capillary plexus of the retina.¹⁶ These changes were verified in animal experiments.^17–19 In contrast, Çavdarlı et al.'s study²⁰ noted a slight increase in VD in the parifovea and perifovea. Patients with hypertension also exhibit thinning retinal thickness and diminished blood flow.^21–23 Hypertension and diabetes mellitus primarily affect retinal health through VEGF-mediated vascular endothelial damaging effects, while inflammatory factors further stimulate VEGF expression.²⁴ 

Alterations in macular thickness are generally considered to be a precursor to macular degeneration, whereas significant vascular loss in the macula is a characteristic feature of nonexudative AMD. Rogala et al.²⁵ have found that macular thickness is reduced significantly in AMD patients, regardless of the presence of drusen, compared with healthy individuals. Furthermore, studies comparing macular layer thicknesses have identified consistently a significant decrease in GCL thickness in AMD patients.^26,27 Ozcaliskan et al.²⁸ observed significantly lower VD in the parafovea superficial capillary plexus in AMD. Toto et al.^29,30 reported that patients with AMD have a lower VD in the superficial vasculature, which is exacerbated further by geographic atrophy owing to structural disorganization and thinning. Notably, Trinh et al.³¹ found that blood perfusion decreases primarily in the temporal quadrant of the retina in AMD, hypothesizing that radial capillaries of the optic disc in the nasal quadrant of the macula compensate for the lack of macular capillaries. This regional specificity aligns with our findings in PTC patients. 

The GCL, owing to its high metabolic demand and synaptic complexity, is particularly susceptible to damage from inflammatory mediators.^32,33 Inflammation compromises the immunosuppressive function of the RPE, resulting in breakdown of the blood–retinal barrier and enhanced permeability.³⁴ In our study, we not only observed significantly elevated NLR and PLR levels in PTC patients, but also identified an inverse correlation between these inflammatory markers and both macular thickness and GCL thickness. Both NLR and PLR have been established as reliable markers associated with the activity and clinical outcome of chronic inflammatory diseases. Pinna et al.³⁵ have found that AMD patients have significantly lower WBC, monocyte, and neutrophil counts than the control group. Ilhan et al.³⁶ have reported significantly higher NLR in patients with AMD than in healthy people, whereas Sengul et al.³⁷ have observed an inverse relationship between visual acuity and NLR as well as PLR. These findings support our conclusion that retinal alterations may be linked to PTC-associated systemic inflammation via blood circulation. Additionally, many studies have suggested that NLR and PLR are linked to the severity of macular disease and can serve as inflammatory markers.^36,38,39 However, Sannan et al.⁴⁰ and Pinna et al.³⁵ argue that these indices may not reliably reflect the retinal changes in AMD owing to the limited correlation between systemic inflammation and localized retinal changes. 

Uveitis is a multifactorial inflammatory eye disease that demonstrates significant alterations in retinal blood flow during the vascular exudative and vascular occlusive phases, characterized by increases and decreases, respectively.^41,42 Studies of quiescent phase posterior or panuveitis patients (excluding those with significant structural abnormalities such as macular scarring, edema, and vascular occlusion) have identified significant decreases in central macular thickness and choroidal thickness in the subcentral sulcus.⁴³ Furthermore, analyses revealed a significant decrease in the superficial capillary density index and vascular branching index, indicating that the superficial and deep vascular plexus of the retina are relatively sparse in patients with resting uveitis. Many studies have found consistent conclusions that patients with uveitis exhibit decreased choroidal thickness, whether compared with myopic or healthy populations.^44–47 The inflammatory response causes changes in the permeability of the vessel wall, leading to inflammatory proteins and inflammatory cells accumulating throughout the body toward the inflamed area.⁴⁸ Consequently, the active phase of uveitis induces changes in photoreceptors, the RPE, and choroidal capillaries, and when patients with active posterior or total uveitis enter a state of quiescence, this transition is accompanied by retinal and choroidal atrophy, along with decreased vascular density.⁴³ In our study, patients with PTC exhibiting invasive tumor features demonstrated increased retinal parameters, suggesting pathological changes analogous to those observed during active uveitis. These elevated retinal parameters likely reflect active proliferation and enhanced metabolism, with the most pronounced increases occurring in the nasal retinal thickness. The increased blood flow signals to invasive and metastatic cancerous tissues may contribute to these changes,⁴⁹ and the predominant changes in the nasal quadrant are consistent with its abundant vascular supply. 

Decreases in retinal thickness and blood flow have been observed in neurological disorders, autoimmune disorders, cerebrovascular disease, and respiratory disease,^50,51 including Parkinson's disease,⁵² Alzheimer's disease,^{53 54} obstructive sleep apnea syndrome,⁵⁵ schizophrenia spectrum disorders,⁵⁶ systemic lupus erythematosus (SLE),^57,58 kidney disease,^59,60 preeclampsia,⁶¹ coronary artery disease,^62,63 carotid artery stenosis,⁶⁴ and COVID-19,⁶⁵ among other systemic conditions. Among these, SLE is a systemic connective tissue inflammation that can progress to lupus retinopathy. A meta-analysis of 13 studies demonstrated reduced VD in both the superficial and deep capillary plexuses in SLE patients.⁶⁶ Notably, retinal microvascular changes often precede renal damage, yet persistent reductions in retinal microvascular blood flow are observed when the renal microvasculature is affected.⁶⁷ The mechanisms underlying SLE-related retinal changes include immune complex deposition in retinal vessels, resulting in endothelial inflammation and microangiopathy, as well as retinal nerve fiber layer (RNFL) and GCL atrophy secondary to chronic hypoxia from impaired perfusion. Analogous pathways may be operative in PTC, we observed a weak positive influence between NK cells, CD4⁺ T cells, and the CD4⁺ T/CD8⁺ T ratio and retinal parameters. This may be because T cell infiltration into retinal degenerative areas and the production of retinal autoantibodies^68,69 may indicate active immune activity contributing to the disruption of retinal structure and blood flow. Eandi et al.⁷⁰ and Sennlaub et al.⁷¹ have observed immune-associated damage in the retinas of patients with AMD, with increased levels of proinflammatory factors detected in serum and aqueous humor. Furthermore, either intraperitoneal injection or parabulbar injection of proinflammatory agents has been shown to induce retinal inflammation,⁷² increasing vitreous inflammatory cell density and subretinal accumulation of activated microglia.^73,74 Penfold et al.⁷⁵ demonstrated autoantibodies against retinal antigens in AMD patients, Johnson et al.⁷⁶ identified specific antibodies against astrocytes in AMD sera, and Żuber-Łaskawiec et al.⁷⁷ observed increased serum levels of antiretinal antibodies in patients with geographic atrophy, suggesting that these antibodies affect retinal capillary permeability. Antigen–antibody immune complexes have also been found in mice with activated microglia and extensive myeloid-cell recruitment in the retina, implicating these complexes in the pathogenesis of macular degeneration.⁷⁸ Moreover, elevated levels of TPOAb, TgAb, and Tg were observed in our PTC group, which are often indicate poor prognosis and high recurrence risk.⁷⁹ However, our study revealed a weakly positive correlation between Tg and retinal thickness. The similarities and differences between antibodies involved in thyroid and retinal require further exploration. 

In summary, it is widely recognized that changes in retinal blood flow serve as early biomarkers for ocular diseases such as AMD, diabetic retinopathy, and uveitis. Therefore, we hypothesize that quantifiable alterations in macular thickness and perfusion parameters may represent a novel ocular biomarker for monitoring systemic inflammatory burden. However, the pathology of microvascular changes involves highly complex mechanisms. Factors such as medications, medical history, and genetics should be extensively studied in the context of retinal alterations to establish a clear link between PTC and retinal changes. 

This study represents an advancement in the field. To our knowledge, it is the first study to explore retinal features in thyroid cancer quantitatively, analyzing the effect of PTC on macular disease from a perspective of inflammatory response. The OCT and OCTA technologies provide valuable insights into retinal and choroidal changes and facilitate studies on fundus lesions in patients with systemic diseases. However, this study has limitations. First, although numerous studies have demonstrated that thyroid hormones affect the retina,^80,81 and our study has also identified limited correlations between T₃, T₄, thyroid-stimulating hormone levels, and retinal thickness, we did not observe significant differences in hormone levels between the PTC group and HCs. Second, our study was limited by the absence of acute phase hematological inflammatory markers, such as procalcitonin and C-reactive protein, which could have provided additional insights. Third, the inflammatory impact of PTC could be evaluated more comprehensively if the retinal parameters and serological parameters of patients with PTC are compared directly with the variability of uveitis or other inflammatory diseases. Therefore, future prospective studies with larger sample sizes and more comprehensive information on systemic and inflammatory diseases compared with HC are essential to better understand retinal structural and microcirculatory changes. 

Conclusions

A trend toward thinner macula and decreased retinal blood flow was observed in PTC patients, with these changes being more pronounced in the temporal quadrant and perifovea of the macula. The observed correlations between PTC and retinal changes, including a weak inverse association with inflammatory markers and a weak positive association with immune cells and aggressive carcinoma signs. Macular blood flow may serve as a surrogate marker for systemic inflammation in the context of cancer. However, it is important to note that these associations may reflect broader systemic effects rather than direct thyroid-specific mechanisms. Confounding factors such as medications, comorbidities, or other inflammatory conditions could also play a role. Ophthalmologists, thyroid surgeons, and primary healthcare providers should be more aware of the impact of PTC on the eye. 

Acknowledgments

The authors thank Jing Liu and Shu-Jing Li of the Thyroid Department for their assistance with patient referral. 

Supported by the National Natural Science Foundation of China (No. 82372009), Shanxi Province Higher Education “Billion Project” Science and Technology Guidance Project (No. BYYX004), Key Research and Development (R&D) Projects of Shanxi Province (No. 201803D31095), Research Project for Returned Scholars Supported by Shanxi Provincial Human Resources and Social Security Department (No. 2019-91-4), Research Projects for Returned Scholars Supported by Shanxi Provincial Department of Finance (No. 2021-163), and Research Project Supported by Shanxi Scholarship Council of China (No. 2024-147). 

Disclosure: Y.-J. Li, None; Z.-Y. Zhang, None; T.-Y. Liu, None; H.-Y. Liu, None; L.-L. Shi, None; R.-X. Cao, None; C.-R. Lv, None; Z.-H. Zhang, None