Abstract
We previously reported that a high HbA1c level 3 months before vitrectomy for vitreous hemorrhage or a large preoperative decrease in the HbA1c level over 3 months tended to increase the risk of rebleeding in diabetic retinopathy patients evaluated between 2010 and 2014. Here, we aimed to confirm these results with an extended study period and an increased number of operated eyes. This study included 121 diabetic patients who were admitted to Osaka University Hospital between 2010 and 2019 and who underwent vitrectomy for vitreous hemorrhage. Binomial logistic regression analysis was performed with the presence of postoperative bleeding as the outcome. The present study showed that the duration of the operation was associated with rebleeding (odds ratio = 1.02, p = 0.0016). A high HbA1c level just before vitrectomy tended to be associated with the bleeding (odds ratio = 1.27, p = 0.05), while preoperative HbA1c changes were not associated with rebleeding. The results of this study suggest that a high preoperative HbA1c level just before vitrectomy, not a decrease in HbA1c levels, in addition to the duration of the operation may increase the risk of postoperative bleeding after vitrectomy in diabetic retinopathy patients.
1. Introduction
Postoperative bleeding is a significant complication following vitrectomy for vitreous hemorrhage in patients with diabetic retinopathy. Moreover, such bleeding delays visual recovery and can make further treatment of retinopathy difficult if the view of the fundus is significantly obscured [1].
It is generally said that rapid amelioration of blood glucose levels could trigger bleeding in diabetic retinopathy patients [2-9]. However, there are no clinical studies that clearly show how rapid blood glucose reduction increases the risk of postoperative bleeding after vitrectomy. Therefore, physicians do not have a clear indicator of preoperative blood glucose control.
In our previous report, we retrospectively examined patients who underwent vitrectomy for vitreous hemorrhage and were admitted to Osaka University Hospital between April 2010 and March 2014 [10]. We found that duration of operation, fasting blood glucose level just before vitrectomy, treatment with antihypertensive drugs and no treatment with antiplatelet drugs were risk factors for postoperative bleeding after vitrectomy. In addition, we found that a trend toward a decrease in the HbA1c level during the preoperative 3 months was associated with postoperative bleeding in 38 eyes [10]. Our previous study was limited by the relatively small number of operated eyes; therefore, further studies with an increased number of operated eyes might be needed to confirm those results.
Here, we conducted a retrospective clinical study to clarify the associations between glycemic control parameters, including the preoperative decrease in HbA1c level, and postoperative bleeding, with an extension of the study period and an increased number of operated eyes.
2. Materials and Methods
2.1 Study population
The subjects of this retrospective study were patients with diabetes (both type 1 and type 2 diabetes mellitus) who were admitted to the Department of Ophthalmology, Osaka University Hospital, Osaka, Japan, between April 2010 and March 2019; who underwent pars plana vitrectomy (PPV) for vitreous hemorrhage associated with proliferative diabetic retinopathy; and whose preoperative HbA1c levels were recorded several times. The diagnostic criteria for diabetes were as follows: fasting blood glucose level ≥126 mg/dL, casual blood glucose level ≥200 mg/dL and/or HbA1c level ≥6.5%. In addition, some patients were diagnosed with diabetes based on the use of antidiabetic medications.
We excluded patients who were diagnosed with glaucoma or detachment of the retina at the time of vitreous hemorrhage. We also excluded patients who underwent PPV performed by surgeons with less than 10 years of experience or who were on hemodialysis due to chronic renal failure. This study was approved by the Institutional Ethics Review Board of Osaka University Hospital (approval number: 20112) and was carried out in accordance with the principles of the Helsinki Declaration. The study was announced to the public on the website of our department at Osaka University Hospital, and all patients were allowed to participate or refuse to participate in the study.
2.2 Patient information
The data for this study were obtained from the medical records of patients kept at Osaka University Hospital. For the preoperative variables, we collected the following data: age, sex, body mass index, systolic and diastolic blood pressure, estimated glomerular filtration rate, fasting blood glucose level just before PPV (preoperative FBS), hemoglobin A1c (HbA1c) level just before PPV (preoperative HbA1c), HbA1c level at approximately 1, 2 or 3 months before PPV, and a decrease in HbA1c before PPV calculated by the following: the difference between the oldest HbA1c value (within three months) and the just before PPV divided by the number of months (%/month, mmol/mol/month). We also collected data regarding diabetes management (diet only, use of oral hypoglycemic agents or use of insulin), history of cardiovascular disease and treatment with antihypertensive or antiplatelet drugs. The operative variables that were collected included the duration of PPV, with or without a concomitant cataract operation, and the number of laser photocoagulation shots during PPV. For postoperative variables, we investigated the incidence of hypoglycemia (below 70 evaluated by self-measurement of blood glucose) within 1 week after surgery.
2.3 Evaluation of postoperative bleeding
For the outcome events, postoperative bleeding after vitrectomy was evaluated at 1, 4, 12, 26 and 52 weeks after surgery. Additionally, the severity of postoperative bleeding was classified as follows according to commonly used classification of vitreous hemorrhage: grade 1, clearly visible optic disk and retinal blood vessels on ophthalmoscopy; grade 2, partial visibility of retinal blood vessels; grade 3, poorly demarcated or ill-defined optic disk; and grade 4, no clear optic disk [11].
Statistical analysis
Binomial logistic regression analysis was carried out. The outcome was defined as the presence or absence of bleeding within 52 weeks after surgery. The data encompassed repeated observations (right and left eyes) for the same patient. The correlation between the duration of PPV and the log transformed number of laser photocoagulation shots during PPV were evaluated by Spearman’s rank correlation coefficient. We conducted statistical analyses using R version 4.2.2 (https://www.r-project.org/). A p value <0.05 was considered to indicate statistical significance.
3. Results
3.1 Patient background
In total, 121 patients (9 type 1 diabetes patients and 112 type 2 diabetes patients) were enrolled (70 males and 51 females). The clinical characteristics of the subjects are shown in Table 1. The median age was 64.0 years. Regarding glycemic control, the median HbA1c level just before PPV (preoperative HbA1c) was 7.4% (57 mmol/mol), and the median 1 month, 2 months and 3 months pre-PPV HbA1c levels were 7.6% (60 mmol/mol), 7.5% (58 mmol/mol) and 7.2% (55 mmol/mol), respectively. The median decrease in HbA1c before surgery calculated by the difference between the oldest HbA1c value (within a three-month period) and the just before PPV divided by the number of months were 0.1%/month (1.1 mmol/mol/month).
Table 1 The clinical characteristics of all patients in the study as well as patients with and without postoperative bleeding
|
Total |
|
Bleeding |
No bleeding |
p value |
| Sex (male/female) |
70/51 |
|
17/13 |
53/38 |
0.92 |
| Age (years) |
64.0 (52.0, 71.0) |
|
52.0 (47.0, 63.0) |
67.0 (58.0, 74.0) |
<0.001 |
| BMI (kg/m2) |
22.8 (21.1, 25.4) |
n = 107 |
22.9 (20.9, 24.1) |
22.9 (21.1, 25.9) |
0.62 |
| Systolic blood pressure (mmHg) |
126 ± 20 |
n = 112 |
126 ± 20 |
128 ± 23 |
0.99 |
| Diastolic blood pressure (mmHg) |
71 ± 12 |
n = 110 |
76 ± 12 |
69 ± 10 |
0.046 |
| eGFR (mL/min/1.73m2) |
56.8 ± 24.5 |
|
49 ± 30 |
58 ± 22 |
0.69 |
| HbA1c just before operation (preoperative HbA1c) (%, mmol/mol) |
7.4 (6.5, 8.2)
57 (48, 66) |
|
7.7 (6.6, 8.8)
60 (48, 73) |
7.3 (6.6, 7.9)
56 (49, 63) |
0.18 |
| HbA1c 3 months before operation (%, mmol/mol) |
7.2 (6.5, 8.1)
55 (48, 65) |
n = 69 |
7.3 (6.4, 7.9)
56 (47, 63) |
7.2 (6.6, 8.2)
55 (49, 66) |
0.86 |
| HbA1c 2 months before operation (%, mmol/mol) |
7.5 (6.6, 8.6)
58 (48, 70) |
n = 52 |
7.5 (6.7, 8.4)
58 (50, 69) |
7.5 (6.6, 8.5)
58 (48, 69) |
0.90 |
| HbA1c 1 month before operation (%, mmol/mol) |
7.6 (6.5, 8.3)
60 (48, 67) |
n = 60 |
8.5 (7.9, 10.2)
69 (62, 88) |
7.4 (6.5, 7.9)
57 (47, 63) |
0.0036 |
| Decrease in HbA1c (%/month, mmol/mol/month) |
0.1 (–0.1, 0.27)
1.1 (–1.1, 2.9) |
|
0.2 (–0.13, 0.4)
2.2 (–1.4, 4.4) |
0.067 (–0.1, 0.24)
0.73 (–1.1, 2.6) |
0.53 |
| Preoperative fasting blood glucose (Preoperative FBG) (mg/dL) |
134 (108, 172) |
n = 98 |
140 (117, 176) |
131 (106, 171) |
0.57 |
| The incidence of hypoglycemia within 1 week after surgery (0/1 or 2/3~) |
98/9/1 |
n = 108 |
27/1/1 |
70/4/4 |
0.91 |
| Diabetes management (diet/oral hypoglycemic agents/insulin) |
24/30/57 |
n = 118 |
6/5/17 |
18/25/40 |
0.57 |
| History of cardiovascular diseases (yes/no) |
35/83 |
n = 118 |
6/20 |
29/60 |
0.27 |
| Treatment with antiplatelet drugs (yes/no) |
32/86 |
n = 118 |
6/20 |
26/63 |
0.43 |
| Treatment with antihypertensive drugs (yes/no) |
91/27 |
n = 118 |
24/4 |
67/22 |
0.25 |
| Number of laser photocoagulation shots |
869 (433, 1414) |
n = 118 |
1192 (795, 1762) |
840 (331, 1375) |
0.015 |
| Duration of the operation (min) |
60 (44, 100) |
n = 119 |
113 (70, 129) |
55 (40, 77) |
<0.001 |
| Concomitant cataract surgeries (yes/no) |
71/50 |
|
18/12 |
53/37 |
0.92 |
The data are presented as the mean ± standard deviation, the median (first quartile, third quartile) or the number of participants.
HbA1c, hemoglobin A1c; BMI, body mass index; eGFR, estimated glomerular filtration rate
Regarding the incidence of hypoglycemia within 1 week after surgery, 98 patients had not experienced hypoglycemia, 9 patients had one or two episodes of hypoglycemia, and 1 patient had more than three episodes of hypoglycemia.
Antidiabetic Medication
Regarding diabetes management, 24 patients were on a diet only, 30 patients were using oral hypoglycemic agents alone, and 57 patients were treated with insulin in combination with or without oral hypoglycemic agents. In our previous report (from 2010 to 2014), the hypoglycemic agents used were as follows: 24% sulfonylurea (SU), 13% thiazolidinedione (TZD), 14% dipeptidyl peptidase-IV (DPP-IV) inhibitor, 11% α-glucosidase inhibitor (αGI), 10% biguanide (BG) and 7% glinide [10]. In the additional subjects from 2014 to 2019, the prescription regimens were as follows: 12% SU, 6% TZD, 24% DPP-IV inhibitor, 9% αGI, 9% metformin, 4% glinide and 3% sodium–glucose cotransporter 2 (SGLT2) inhibitor. Among the additional patients, 48% received insulin therapy, which was not different from the 42% reported previously [10]. Table 1 provides details about the patients’ cardiovascular diseases and their use of antihypertensive and antiplatelet drugs.
3.2 Operative variable
The operative variables are also listed in Table 1. The median number of laser photocoagulation shots taken during surgery was 869. The duration of the operation was 60 min. Furthermore, 71 patients had concomitant cataract surgeries. The duration of PPV was significantly associated with the log transformed number of laser photocoagulation shots during PPV (p < 0.001).
3.3 Association between postoperative bleeding and preoperative HbA1c or other factors
Postoperative bleeding within 52 weeks after vitrectomy occurred in 30 patients. The cases of postoperative bleeding often occurred within a week. The grade of postoperative bleeding is shown in Table 2. When the participants were divided into two groups according to the presence or absence of postoperative bleeding, the duration of the operation was greater, the age was younger, the HbA1c level 1 month before surgery was higher, and the number of laser photocoagulation shots was also greater in the group with postoperative bleeding than in the group without postoperative bleeding. These results are shown in Table 1.
Table 2 The time after vitrectomy and grade of postoperative bleeding in each patient
| Patient No. |
1 week |
1 month
(4 weeks) |
3 months
(12 weeks) |
6 months
(26 weeks) |
1 year
(52 weeks) |
| 1 |
3 |
0 |
0 |
0 |
ND |
| 2 |
0 |
0 |
1 |
3 |
0 |
| 3 |
0 |
0 |
1 |
0 |
ND |
| 4 |
0 |
0 |
3 |
0 |
0 |
| 5 |
1 |
0 |
0 |
0 |
ND |
| 6 |
0 |
4 |
0 |
0 |
0 |
| 7 |
1 |
0 |
0 |
0 |
0 |
| 8 |
1 |
0 |
0 |
0 |
0 |
| 9 |
0 |
4 |
4 |
4 |
0 |
| 10 |
1 |
0 |
0 |
0 |
0 |
| 11 |
1 |
4 |
0 |
1 |
1 |
| 12 |
4 |
2 |
0 |
0 |
0 |
| 13 |
1 |
1 |
1 |
1 |
1 |
| 14 |
1 |
0 |
0 |
0 |
0 |
| 15 |
2 |
1 |
1 |
0 |
0 |
| 16 |
4 |
1 |
0 |
0 |
0 |
| 17 |
1 |
0 |
0 |
0 |
0 |
| 18 |
1 |
0 |
0 |
0 |
0 |
| 19 |
4 |
0 |
0 |
0 |
1 |
| 20 |
2 |
0 |
0 |
4 |
0 |
| 21 |
4 |
0 |
0 |
4 |
0 |
| 22 |
0 |
0 |
0 |
4 |
0 |
| 23 |
1 |
0 |
ND |
ND |
0 |
| 24 |
4 |
2 |
0 |
0 |
0 |
| 25 |
1 |
0 |
0 |
0 |
ND |
| 26 |
3 |
0 |
0 |
4 |
0 |
| 27 |
4 |
4 |
0 |
0 |
ND |
| 28 |
4 |
4 |
0 |
0 |
ND |
| 29 |
2 |
0 |
0 |
0 |
0 |
| 30 |
4 |
4 |
4 |
3 |
4 |
0: no bleeding; 1: optic disk and retinal blood vessels are observed clearly; 2: retinal blood vessels are observed to some extent; 3: optic disk is observed, but not clearly; 4: optic disk is not observed; ND: not determined
A binomial logistic regression analysis was applied to determine the predictors of postoperative bleeding (Table 3). We analyzed the distinctive factors described in Table 3 after adjustment for age. Among these factors, the duration of the operation was identified as a significant predictor of postoperative bleeding. The decrease in HbA1c before surgery was not significant. A high HbA1c level just before vitrectomy tended to be associated with the bleeding.
Table 3 Binomial logistic regression analysis for predicting postoperative bleeding
|
Odds ratio |
95% Confidence interval |
p value |
| Systolic blood pressure (mmHg) |
0.99 |
0.97–1.02 |
0.57 |
| Diastolic blood pressure (mmHg) |
1.00 |
0.96–1.04 |
0.99 |
| Preoperative FBG (mg/dL) |
1.00 |
0.99–1.01 |
0.88 |
| HbA1c just before operation (preoperative HbA1c) (%) |
1.27 |
1.00–1.62 |
0.050 |
| Decrease in HbA1c (%/month) |
0.86 |
0.33–2.25 |
0.76 |
| The incidence of hypoglycemia within 1 week after surgery |
0.92 |
0.35–2.46 |
0.88 |
| Treatment with antiplatelet drugs, reference = ‘no’ |
1.00 |
0.33–3.01 |
0.99 |
| Treatment with antihypertensive drugs, reference = ‘no’ |
1.97 |
0.58–6.72 |
0.28 |
| Log (The number of laser photocoagulation shots) |
1.60 |
0.79–3.26 |
0.19 |
| Duration of the operation (min) |
1.02 |
1.01–1.04 |
0.0016 |
| Concomitant cataract surgeries, reference = ‘no’ |
0.86 |
0.34–2.16 |
0.74 |
All the data were analyzed after adjusting for age.
4. Discussion
In this retrospective clinical study, we sought to clarify the associations between glycemic parameters and postoperative bleeding with an extension of the study period and an increase in the number of operated eyes. We expected that a large preoperative decrease in HbA1c might be significantly associated with an increased risk of bleeding, however, we could not show such results. On the other hand, a high HbA1c level just before vitrectomy tended to be associated with the bleeding.
The reason why the preoperative decrease in HbA1c was not associated with bleeding in the present study might be that the median preoperative HbA1c level in the present study was not very high (7.4%) (57 mmol/mol), unlike that in previous reports [2, 3, 5-7]. This might have caused the median decrease in preoperative HbA1c level to be quite small in the present study (0.1%/month) (1.1 mmol/mol/month). However, further studies of patients with poorer glycemic control or patients who exhibit a greater decrease in preoperative HbA1c levels are needed.
High HbA1c levels are reportedly associated with high serum VEGF levels [12]; therefore, the risk of postoperative bleeding due to hyperglycemia may be related to the neovascularization of fibrovascular membranes [13]. It is also possible that hyperglycemia induces inflammation, resulting in capillary dysfunction [14]. Indeed, some previous reports have shown that bleeding after vitrectomy is more likely to occur in patients with high preoperative HbA1c levels [15] and hyperglycemia [16], which was partly consistent with the results of our present study.
The reason why the duration of the operation was associated with postoperative bleeding may be as follows: The operative time included the intraoperative retinal photocoagulation irradiation time and hemostasis time. Indeed, the duration of PPV was closely associated with the number of laser photocoagulation shots in this study. Therefore, a longer operation time may have reflected the difficulty of the procedure and the severity of the retinopathy; thus, a longer operation time may be a predictor of rebleeding after vitrectomy.
This study has several limitations. First, the median HbA1c levels of the patients were not so high, and the results shown in this study are not applicable to all diabetes patients. Second, the glycemic control 1 month before surgery was worse and the age was younger in the group with postoperative bleeding, which implies that this group may include more younger-onset, poorly glycemic-controlled diabetes patients, who are thought to be high risk for severe retinopathy [17]. This may have affected the results in this study. Third, the number of included patients and the number of patients with postoperative bleeding were small, and thus, the analyses were adjusted for age only. In addition, we believe that patient surveys based on a prospective analysis of factors, including numerous ophthalmologic factors, are needed. Finally, we could not determine the exact duration of diabetes mellitus cases, which may have had a significant impact on postoperative bleeding after vitrectomy, as this was a retrospective study.
In conclusion, the results of this study suggest that during vitrectomy for vitreous hemorrhage in diabetic retinopathy patients, a high HbA1c level just before operation may increase the risk of postoperative bleeding. Improving blood glucose levels before vitrectomy may be desirable for diabetes patients with poor glycemic control.
Acknowledgements
We thank Satomi Okamura from the Data Coordinating Center, Department of Medical Innovation, Osaka University Hospital, Suita, Japan, for advice on statistical analysis.
Declaration of competing interests
The authors declare no conflicts of interest.
Ethics approval
This study was approved by the Institutional Ethics Review Board of Osaka University Hospital (approval number: 20112) and was carried out in accordance with the principles of the Helsinki Declaration. The study was announced to the public on the website of our department at Osaka University Hospital, and all patients were allowed to participate or refuse to participate in the study.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Author contributions
S.M. collected the data, analyzed the data and wrote the manuscript. M.K. collected the data, analyzed the data and contributed to the discussion. S.S., C.I., S.F., Y.H., M.Y.B., Y.F., T.K., A.T., T.N. and K.F. contributed to the discussion. J.K., H.S., K.N. and I.S. contributed to the discussion and reviewed/edited the manuscript. J.K. is the guarantor of this work and as such had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Availability of data and material
All data relevant to the present study are included in this article. The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.
References
- 1 Smith JM, Steel DH (2015) Anti-vascular endothelial growth factor for prevention of postoperative vitreous cavity haemorrhage after vitrectomy for proliferative diabetic retinopathy. Cochrane Database Syst Rev 2015: CD008214.
- 2 Brinchmann-Hansen O, Dahl-Jorgensen K, Hanssen KF, Sandvik L (1985) Effects of intensified insulin treatment on various lesions of diabetic retinopathy. Am J Ophthalmol 100: 644–653.
- 3 The Kroc Collaborative Study Group (1984) Blood glucose control and the evolution of diabetic retinopathy and albuminuria. A preliminary multicenter trial. N Engl J Med 311: 365–372.
- 4 Funatsu H, Yamashita H, Ohashi Y, Ishigaki T (1992) Effect of rapid glycemic control on progression of diabetic retinopathy. Jpn J Ophthalmol 36: 356–367.
- 5 The Diabetes Control and Complications Trial Research Group (1998) Early worsening of diabetic retinopathy in the diabetes control and complications trial. Arch Ophthalmol 116: 874–886.
- 6 Shurter A, Genter P, Ouyang D, Ipp E (2013) Euglycemic progression: worsening of diabetic retinopathy in poorly controlled type 2 diabetes in minorities. Diabetes Res Clin Pract 100: 362–367.
- 7 Vilsbøll T, Bain SC, Leiter LA, Lingvay I, Matthews D, et al. (2018) Semaglutide, reduction in glycated haemoglobin and the risk of diabetic retinopathy. Diabetes Obes Metab 20: 889–897.
- 8 Testa MA, Puklin JE, Sherwin RS, Simonson DC; The Kroc Collaborative Study Group (1985) Clinical predictors of retinopathy and its progression in patients with type I diabetes during CSII or conventional insulin treatment. Dibetes 34(Suppl 3): 61–68.
- 9 Bethel MA, Diaz R, Castellana N, Bhattacharya I, Gerstein HC, et al. (2021) HbA1c change and diabetic retinopathy during GLP-1 receptor agonist cardiovascular outcome trials: a meta-analysis and meta-regression. Diabetes Care 44: 290–296.
- 10 Motoda S, Shiraki N, Ishihara T, Sakaguchi H, Kabata D, et al. (2018) Predictors of postoperative bleeding after vitrectomy for vitreous hemorrhage in patients with diabetic retinopathy. J Diabetes Investig 9: 940–945.
- 11 Ahmadieh H, Shoeibi N, Entezari M, Monshizadeh R (2009) Intravitreal bevacizumab for prevention of early postvitrectomy hemorrhage in diabetic patients. Ophthalmology 116: 1943–1948.
- 12 Zehetner C, Kirchmair R, Kralinger M, Kieselbach G (2013) Correlation of vascular endothelial growth factor plasma levels and glycemic control in patients with diabetic retinopathy. Acta Ophthalmol 91: 470–473.
- 13 Aiello LP, Avery RL, Arrigg PG, Keyt BA, Jampel HD, et al. (1994) Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N Engl J Med 331: 1480–1487.
- 14 Wang L, Wang J, Fang J, Zhou H, Liu X, et al. (2015) High glucose induces and activates Toll-like receptor 4 in endothelial cells of diabetic retinopathy. Diabetol Metab Syndr 7: 89.
- 15 Sato T, Tsuboi K, Nakashima H, Emi K (2017) Characteristics of cases with postoperative vitreous hemorrhage after 25-gauge vitrectomy for repair of proliferative diabetic retinopathy. Graefes Arch Clin Exp Ophthalmol 255: 665–671.
- 16 Mahalingam P, Topiwalla TT, Ganesan G (2018) Vitreous rebleed following sutureless vitrectomy: incidence and risk factors. Indian J Ophthalmol 66: 558–561.
- 17 Perais J, Agarwal R, Evans JR, Loveman E, Colquitt JL, et al. (2023) Prognostic factors for the development and progression of proliferative diabetic retinopathy in people with diabetic retinopathy. Cochrane Database Syst Rev 2: CD013775.
