Quantitative assessment of early Type 2 diabetic cataracts using T1,T2-mapping techniques

Abstract

Objective:

The purpose of this study was to explore the feasibility of T1 and T2-mapping techniques in evaluating early Type 2 diabetic cataracts.

Methods:

Totally, 28 patients with Type 2 diabetes was prospectively collected, and 28 non-diabetic patients were collected as control group. All patients included had ophthalmological exploration and all patients underwent orbital MRI examination with T1 and T2-mapping on a Siemens-Skyra 3.0T scanner. T1 and T2 values of the lens nucleus were measured by region of interest (ROI) method based on Siemens-Syngo workstation. Two sample t-test was used to analyze the differences between groups. Pearson correlations were calculated between relaxation time (T1, T2) and clinical variables, such as fasting glucose, glycosylated hemoglobin etc. p < 0.01 was used to determine statistical significance.

Results:

In Type 2 diabetes group, the T1 value was 626.7 ± 56.8, T2 value was 29.4 ± 5.6. In non-diabetic group, the T1 value was 581.6 ± 64.7, T2 value was 24.8 ± 8.6. The T1 and T2 values of the lens in diabetic group were significantly higher than those in control group (p < 0.01, T1 value: 626.7 ± 56.8 vs 581.6 ± 64.7; T2 value: 29.4 ± 5.6 vs 24.8 ± 8.6). The T1 and T2 values of lens in diabetic patients were significantly correlated with glycosylated hemoglobin (HbA1c), and the correlation coefficients were 0.502 and 0.396, respectively.

Conclusion:

T1 and T2-mapping technique can sensitively reflect the alterative relaxation time of lens in diabetic patients. This technique can find abnormal changes earlier than slit lamp, and may be effective diagnostic methods for early lens disease.

Advances in knowledge:

T1 and T2-mapping techniques may be effective diagnostic methods for early lens disease, which can detect abnormal changes earlier than slit lamp examination.

Introduction

Type 2 diabetes (T2D) is a risk factor for cataract development. With T2D prevalence increasing, the burden of cataract-associated vision loss will also increase.¹ Despite its high incidence cataract is still poorly understood, which may be attributed to the fact that it is generally detected at an advanced stage. As cataract is a disabled disease with a negative prognosis and a high socioeconomic impact, it is crucial to develop techniques for its early detection.

Before structural changes manifest, the lens tissue is subjected to biochemical alterations, including an accumulation of advanced glycation endproducts (AGE), sorbitol and a concomitant increase in water content.² These early changes cannot be detected using "classical" methods such as slit lamp examination.

Previous studies have shown that the water content in diabetes affected lens may increase. For the early detection of lens disease using MR methods, the water content may be the biomarker of choice, as it requires neither a lengthy study time nor the administration of contrast agents. T1 and T2-mapping technology can detect tissue relaxation value in vivo and reflect tissue biochemical changes,^3,4 which was recently demonstrated to be very sensitive to the water content. This technology has been widely used in clinic to assess joint diseases.^5,6 At present, this technology has not been applied to evaluate human lens disease in vivo. According to the microanatomy of lens, the relaxation value of lens is relatively stable under normal condition. However, under pathological conditions, with the change of biochemical compositions, tissue relaxation time will change. How will the lens relaxation time change in diabetic patients?

To answer the above questions, this study was designed to use T1 and T2-mapping techniques to detect the changes in relaxation time of lens in diabetic patients. And also to explore the feasibility of this technology in evaluating early lens disease in diabetes. The optimal goal of MRI of the lens may be to follow-up of disease progression and its response to treatment.

Methods and materials

Subjects

Totally, 28 patients with Type 2 diabetes were prospectively collected, They were diagnosed with diabetes for more than 3 years. And 28 non-diabetic patients were collected as control group. After approval by the local ethics committee, informed consent was obtained from all participants. Exclusion criteria¹: slit lamp examination with lens opacity²; glaucoma patients³; high blood lipids, hyperthyroidism and other metabolic diseases⁴; had a history of ocular trauma. Venous blood was collected from all diabetic patients to measure fasting blood glucose, glycosylated hemoglobin. All subjects received eye slit lamp examination.

MRI image acquisition

We performed all the MRI examination on a 3.0 T scanner (Magnetom Skyra, Siemens Healthineers, Erlangen, Germany) using a 24-channel head coil. All the subjects were asked to take supine position and kept eyes closed meanwhile wearing eyeshade, a sponge pad was used to fix head, cotton balls placed inside the canthus to keep the eyeballs stationary. When the examination started, the patient was asked to keep the eyeball not rotated. The examination protocols included a T1-mapping sequence [repetition time (TR) = 15 ms, echo time (TE) = 2.2 ms, flip angle₁ = 5^°, flip angle₂ = 26^°, field of view 170 × 53 mm, thickness 3.0 mm, number of slices 12], a CPMG-based T2-mapping sequence (TR = 600 ms, TE₁ = 13.8 ms, TE₂ = 27.6, TE₃ = 41.4, TE₄ = 55.2, flip angle 180^°, field of view 170 × 53 mm, thickness 4.0 mm, number of slices 7).

Image analysis

Measurements of MR parameters were performed by two radiologists (with 7 and 15 years’ experience in MRI of head and neck region, respectively) independently. Both of them were blinded to the clinical data of patients. After acquisition of T1 and T2-mapping, T1 and T2-maps were automatically calculated on the scanning workstation. The regions of interest (ROIs) were drawn within largest slice of lens to calculate T1 and T2 value (ROI size 0.04 cm², placed in the central region of the lens). See Figure 1.

Figure 1.

Representative color-coded T2 and T1 maps (a, b). A 45-year-old female, diagnosed with diabetes for 5 years. (a) T2-maps; (b) T1-maps. The ROIs were drawn within largest slice of lens to calculate T1 and T2 value (ROI size 0.04 cm², placed in the central region of the lens). ROI, region of interest

Statistical analysis

Continuous data were shown as mean ± standard deviation. Two independent samples t-test was used to compare T1 and T2 value between diabetes group and non-diabetic group. The Pearson χ² test was performed to assess the differences in gender between the two groups. Pearson correlations were calculated between relaxation value (T1, T2) and clinical variables (fasting glucose, glycosylated hemoglobin and duration of disease) in patients with diabetes. Intraclass correlation coefficient (ICC) was used to evaluate the reproducibility of T2 and T1 values measurements. All statistical analyses were performed using SPSS 18.0 software (SPSS Inc., Chicago, IL). p < 0.01 for statistical analyses indicated a statistical significance.

Results

Population

According to the inclusion criteria, there were 28 cases of Type 2 diabetes mellitus (17 males with mean age of 49.1 ± 9.2; 11 females with mean age of 50.0 ± 8.9). Totally, 28 cases were in the control group (15 males with mean age of 53.5 ± 5.5; 13 female with mean age of 45.6 ± 98, p > 0.01). There was no significant difference in age and sex between the groups (p > 0.01). Epidemiologic and pathologic information of all 28 patients and 28 controls are shown in Table 1.

Table 1.

HbA1, FPG, age, gender, and duration of disease in the study group

	Diabetes	Controls	t/χ	p
Fasting glucose (mmo/l)	6.53 ± 1.65	4.92 ± 0.73	6.708	<0.01
Glycosylated hemoglobin (%)	7.42 ± 1.03	——	——	——
Duration of disease (years)	6.00 ± 2.50	——	——	——
Gender(F/M）	11/17	13/15	0.292	0.589
Age (years)	50.6 ± 9.1	50.0 ± 8.6	0.471	0.638

FPG, fasting plasma glucose.

The inter- and intraobserver consistency test results

The inter- and intraobserver agreement test results are shown in Table 2. In both intra- and interobserver agreements, ICC for measurement of relaxation value was interpreted as substantial agreement.

Table 2.

Inter- and intraobserver consistency test results

	Interobserver (ICC)		Intraobserver (ICC)
	T1	T2	Observer1		Observer 2
			T1	T2	T1	T2
Diabetes	0.980	0.868	0.992	0.852	0.978	0.969
Controls	0.955	0.961	0.999	0.981	0.995	0.956

ICC, intraclass correlation coefficient.

Comparison of lens relaxation time between diabetic group and control group

In Type 2 diabetes group, the T1 value was 626.7 ± 56.8 ms, T2 value was 29.4 ± 5.6 ms. In non-diabetic group, the T1 value was 581.6 ± 64.7 ms, T2 value was 24.8 ± 8.6 ms. The T1 and T2 values of the lens in diabetic group were significantly higher than those in control group (T1 value: 626.7 ± 56.8 vs 581.6 ± 64.7, t = 3.91, p < 0.01; T2 value: 29.4 ± 5.6 vs 24.8 ± 8.6, T = 3.34, p < 0.01), for detailed information, see Figure 2 .

Figure 2.

(a) T1 value in the diabetes group and healthy controls, t = 3.91, p < 0.01. (b) T2 value in the diabetes group and healthy controls, t = 3.34, p < 0.01

Pearson correlation analysis of clinical variables and lens relaxation values

The T1 and T2 values of lens in diabetic patients were significantly correlated with glycosylated hemoglobin values, and the correlation coefficients were 0.502 and 0.396, respectively (Figure 3). Fasting blood glucose, duration of disease had little correlation with relaxation value, p > 0.01.

Figure 3.

(a) Correlation between glycosylated hemoglobin and lens T1 value, the correlation coefficients were 0.502, p < 0.01. (b) Correlation between glycosylated hemoglobin and lens T2 value, the correlation coefficients were 0.396, p < 0.01

Discussion

In order to find an objective monitoring tools for Type 2 diabetic cataracts at early stage, this study have used T1 and T2-mapping techniques to detect the changes of lens relaxation time. It was demonstrated that the decoding of lens disease with T1 and T2-mapping technology is feasible. The results showed that the relaxation time of lens (T1, T2) in diabetic patients were significantly higher than those in non-diabetic patients, which was earlier than the slit-lamp examination. In addition, it was found that the relaxation time was positively correlated with glycosylated hemoglobin.

Relaxation time is a biological MR parameter. Each tissue type exhibits a characteristic range of normal relaxation times at a particular magnetic field strength. At the early stage of most disease, the tissue is subjected to biochemical alterations with a concomitant increase in water content.^7–9 Thus the most important factor, which determines tissue relaxation time at this stage, may be the tissue water content. For this reason, changes in tissue water content have been recommended by some researchers as biomarkers for early diagnosis.^7,9–11 Previous studies have confirmed that the water content of lens increases significantly at the early stage of diabetes, it is generally believed that cataract formation in diabetic lenses has been attributed to polyol-osmotic pressure-generated influx of water.^2,12 T1 and T2-mapping techniques are very sensitive to the changes in water content.^8,9 Our study found that T1 and T2 value of the lens in diabetic patients increased significantly. We speculated that this phenomenon may be related to the increase of water content in lens. The results are consistent with previous opinion about the polyol-osmotic pressure-generated influx of water. In addition, the lens with abnormal relaxation time could not be detected under the slit lamp microscope. It suggests that lens may have abnormal biochemical alterations at an early stage, and slit-lamp examinations are not sensitive to these changes. The T1 and T2-mapping can be used as a sensitive method to detect diabetic-related lens disease at an early stage.

This study also found that the relaxation value of lens was significantly positive correlated with the level of glycated hemoglobin (HbA1c). HbA1c is the best accepted measure of dysglycemia for the past 3 months, which reflects long-term glycemic control. Many studies have found that the level of glycosylated hemoglobin is closely related to the severity of various diabetic complications. Wanget al found that glycemic control were consistently associated with cataract development in Type 2 diabetes mellitus.¹³ Our result suggests that suboptimal glycemic control may be a risk factor for lens disease development. For Type 2 diabetes mellitus with suboptimal glycemic control, T1 and T2-mapping techniques are expected to detect biochemical changes at the early stages of lens disease. In addition, it also can provide imaging evidence for clinical treatment options, as some studies have found that early diabetic lens disease can be reversed by emerging therapeutic technologies.^14–16

Our study also has some limitations. First, neither biochemical nor histological assessment of lens was performed in this study, and therefore the relationship between severity evaluation obtained by our method and actual pathological state obtained by lens is still unclear. Secondly, this study does not involve detection after treatment. Therefore, whether small changes after treatment can be detected need further study. Third, the patients in this study do not include mild cataract patients due to excluding, of patients with lens opacity found by slit lamp examination, so we can not make a conclusions of patients with mild cataract. In the future, we will supplement experimental data to explore the mapping data of different conditions and time course of cataract. Fourth, the samples used in this study are small, although we have obtained meaningful results, large sample studies are still needed to further verify the accuracy of the results.

In conclusion, T1 and T2-mapping technique can sensitively reflect the alterative relaxation time of lens in diabetic patients. This technique can find abnormal changes earlier than slit lamp, and may be an effective diagnostic method for lens disease at early stage.