Diabetic patients with PDR were recruited between 1st June 2004 and 30th June 2015. The mean follow-up period was 8.47 ± 1.17 years [7-21]. Patient demographics are shown in Table 1.
The majority of participants were men [197/69%] and 88 were women [31%]. Approximately one third of eyes with PDR had NVH for both genders. Women had a greater proportion of eyes with SVH and men had double the number of eyes with RVH [p=0.001].
In the present study, 183 eyes with PDR had vitreous haemorrhage [64%], of which 68 eyes [37%] had RVH. The majority of eyes that had RVH re-bled once [61%] and twice [33%]. Sixty-four eyes [85%] had late RVH and 11 eyes [15%] had early RVH.
Study of risk factors
As shown in Table 1, the groups of patients that suffered from NVH and SVH were mainly the non-smokers whereas the group with RVH were mainly smokers [p<0.001]. The mean pack-years of those who smoked in the NVH group was 10.13 ± 18.61, in the SVH group was 25.32 ± 35.94 and in the RVH group 49.34 ± 44.42, this result being statistically significant [p<0.001]. Therefore, there was a greater proportion of smokers in the RVH group who smoked more. In addition, there was a positive correlation between the number of vitreous haemorrhage episodes and the pack-years consumed [r=0.339; p<0.001] [Figure 1]. By contrast, there was no significant difference observed in the vitreous haemorrhage between former smokers and non-smokers.
Patients in the RVH group had longer duration of DM compared with the SVH and NVH groups. The longer the duration, the greater number of vitreous haemorrhage episodes [r=0.128; p=0.031] [Table 1, Figure 1].
The mean of haemoglobin was different among groups: NVH = 13.21 ± 1.70, SVH = 12.60 ± 1.55 and RVH = 11.71 ± 1.89 [p<0,001]. The lower the haemoglobin level the more episodes of vitreous haemorrhage [r= - 0.260; p<0.001] [Figure 1]. In the present study, as shown in Table 2, tobacco consumption, duration of diabetes and haemoglobin level had an independently significant associaton with the presence of recurrent vitreous haemorrhage. The Kaplan-Meier survival analysis estimated that after 25 years of diabetes duration, 40% of smokers had recurrent vitreous haemorrhage while only 14% of former smokers and 10% of non smokers [Log Rank=22.15; p< 0.005 ] [Figure 2].
Regarding the status of the vitreous, there were significant differences among the three groups of patients at the beginning of the study [p<0.001]. In total, 177 eyes [62%] had the vitreous attached to the retina whereas 108 eyes [38%] showed PVD. Twice as many eyes of the patients in the NVH group had the vitreous attached to the retina. Patients in the SVH group had the same proportion of eyes with and without the vitreous attached to the retina whereas three-quarters of the number of eyes of patients in the RVH group had the vitreous attached to the retina. Patients with the vitreous attached to the retina were younger [57.86 ± 11.77] than those with PVD [72.65 ± 9.79] and there was no relationship between the presence of recurrent vitreous haemorrhage and the status of the vitreous (OR= 2.15, 95% CI, 0.93-4.93, P= 0.07).
Patients diagnosed with COPD were more likely to show vitreous re-bleeding. Nevertheless, once this variable was studied through the logistic regression analysis it could not be considered an independent risk factor for RVH because it correlated positively with smoking.
Nearly all patients [95%] who took anticoagulant drugs in our series had SVH or RVH whereas that was the case for only 60% of those patients in the non-anticoagulant group. However, anticoagulants was not an independent risk factor for RVH since it correlated negatively with haemoglobin and positively with pack-years in our series. The proportion of patients that took antiplatelet drugs was similar among the three groups and was not a risk factor for eye re-bleeding.
In the present study, the mean of HbA1c was similar among the three group of patients and we did not observe a correlation between the level of HbA1c and the number of vitreous haemorrhage events. However, men showed a lower mean HbA1c level [8.49 ± 1.70] than women [9.10 ± 1.49] [p=0.004].
We included 40 eyes of T1DM patients in our study. Sixteen eyes had NVH [40%], 11 eyes [27.5%] had SVH and 13 eyes [32.5%] had RVH. Also, 245 eyes came from T2DM patients of whom 86 eyes [35%] did not bleed, 104 eyes [42.5%] had SVH and 55 eyes [22.5%] had RVH. There was no significant difference in terms of vitreous haemorrhage behaviour between T1DM and T2DM patients. Regarding treatment, 30 patients were treated with oral antidiabetic drugs [18.5%] and 135 were treated with insulin [81.5%]. The mean of vitreous haemorrhage episodes observed was similar for both groups of patients.
The BMI was similar among groups at the start of the study and did not correlate with the tendency to show vitreous haemorrhage.
In relation to both the UACR and the eGFR, there were no differences among the three groups at the start of the study and there was no significant tendency to suffer from RVH.
In the present study, 175 eyes with PDR [61,4%] were referred with no previous PRP carried out. Eighty-one eyes [46.3%] did not have VH, 65 eyes [37.1%] had SVH and 29 eyes [16.6%] developed RVH. By constast, 110 eyes [38.6%] had previously undergone PRP of which 21 eyes [19.1%] did not develop vitreous haemorrhage, 50 eyes [45.5%] developed SVH and 39 eyes [35.5%] developed RVH. Therefore, a greater percentage of eyes that had previously undergone PRP, developed SVH and RVH [p<0.001].
Diabetic macular oedema.
We evaluated the presence and type of DME in the three groups of eyes with PDR at the start of the study. Twenty-eight eyes [27.5%] in the NVH group did not have DME, 44 eyes [43,1%] had focal DME and 30 eyes [29,4%] had diffuse DME. In the SVH group, 31 eyes [27%] did not have DME, 52 eyes [45.2%] had focal DME and 32 eyes [27.8%] had diffuse DME. Finally, in the RVH group, 12 eyes [17.6%] did not present with DME, 17 eyes [25%] had focal DME and 39 eyes [57.4%] had diffuse DME. Therefore, eyes with RVH presented with diffuse DME significantly greater percentage than the others [p<0.001].
Study of anti-VEGF.
One hundred and four eyes included in the study [36.5%] had previously received anti-VEGF injections [mostly ranibizumab] for DME. Most eyes received 3 injections [37.5%]. The mean number of vitreous haemorrhages in the group of eyes that received anti-VEGF injections was 0.46 ± 0.823 compared to 0.38 ± 0.755 in those that did not. Therefore, the previous treatment with anti-VEGF therapy, independently of the number of injections, did not influence the tendency to re-bleeding.
Relationship with other diabetes complications.
The existence of systemic complications was recorded in all patients. Approximately one third of patients [38%] with diabetic polineuropathy developed RVH, which is a higher percentage than the others [p<0.001]. Similarly, those patients diagnosed with some type of ischemia in lower limbs [35%] were more prone to show RVH when compared to the others [p<0.001]. In the same way, those patients who had ischemic cardiopathy [25%] were more prone to suffer from RVH [p<0,001]. The rate of occurrence of stroke was similar among groups and did not relate with eye re-bleeding.