23-gauge transconjunctival vitrectomy in eyes with pre-existing functioning filtering blebs

Trabeculectomy and vitrectomy are procedures most often performed in glaucoma and retinal disease. The number of patients with functioning filtering bleb may require PPV for various vitreoretinal disease. Eyes that receive prior trabeculectomy can lead to problems if further retinal diseases requiring vitrectomy. Trabeculectomy has long been selected as an effective surgical method to maintain low IOP for prolonged periods of time. Antimetabolites such as MMC or 5-FU can be used to increase the success rate of surgery. MMC or 5-FU regulates the wound healing process of bleb by suppressing subconjunctival fibrosis that is the major cause of surgical failure and also affects bleb function and morphology [8]. Preservation of conjunctiva after the surgery is important because success of the surgery depends on the condition of bleb after trabeculectomy. Thus, when patients with functioning filtering bleb may be required PPV, it is considerable factor how well the filtering bleb is preserved. In 1976, Kolker and Heathering reported a 50 % loss of filtering bleb in intracapsular cataract extraction carried out after trabeculectomy [9]. However, with development of phacoemulsification, Alpar reported that loss of filtering bleb occurred in only 1 of 7 eyes with phacoemulsification after trabeculectomy [3]. Likewise, when vitrectomy is needed in eyes that receive prior trabeculectomy, it would be more advantageous to use transconjunctival sutureless microincisional PPV than the 20G conventional PPV. Thompson et al. reported that it is difficult to maintain bleb function when 20G conventional PPV is performed after trabeculectomy [4]. Vitreoretinal surgery has recently been developing as minimal invasive technique. Eyes with prior trabeculectomy require microincision vitrectomy for preservation of the filtering bleb. In addition, conjunctival scarring occurs in most cases after 20G conventional PPV, and transconjunctival sutureless microincisional vitrectomy can minimize conjunctival scarring and increase success rate of trabeculectomy later on [10]. There was a case in 2007 that reported successful outcome of 25G PPV in familial amyloid polyneuropathy patients with filtering bleb [11]. Kunikata et al. recently reported the stable results of 25G microincision vitrectomy in eyes that received trabeculectomy [7]. Therefore, the authors evaluated 23G PPV for vitreoretinal disease in eyes with prior trabeculectomy.

In this study, 36 % of the subjects had open angle glaucoma for which trabeculectomy was performed, followed by 27 % with neovascular glaucoma. Kunikata et al. reported 53 % neovascular glaucoma and 13 % open angle glaucoma cases, and Thompson et al., reported 57 % open angle glaucoma and 17 % angle closure glaucoma [4, 7]. In the current study, 4 eyes (36 %) of the 11 subjects had epiretinal membrane and 3 eyes (27 %) of the 11 subjects had proliferative diabetic retinopathy. According to Kunikata et al., 53 % of patients who received 25G PPV in eyes that received trabeculectomy were proliferative diabetic retinopathy patients, and only 1 eye (6.5 %) was from an epiretinal membrane patient [7]. Thompson et al., studied patients with 20G PPV in eyes that received trabeculectomy, of which 57 % had rhegmatogenous retinal detachment and 13 % each had proliferative diabetic retinopathy and epiretinal membrane [4]. The high ratio of epiretinal membrane in our study is likely due to the expansion of epiretinal membrane surgery indication by the development of OCT compared to before. Thompson et al. reported that the use of prior antimetabolite was not correlated with preservation of bleb function [4]. We had statistical difficulty comparing the use and non-use of prior antimetabolite because all cases used antimetabolite. Although the use of prior antimetabolite does not affect preservation of bleb function, measures such as 5-FU injection after PPV are believed to maintain bleb function. Nonetheless, since the possibility of bleb failure is low in transconjunctival sutureless microincisional vitrectomy as in 23G PPV, 5-FU injection can be attempted immediately after PPV when bleb failure is a risk.

BCVA was improved in all cases after 23G PPV, which implies that the status of bleb does not affect the success rate of vitrectomy. Thompson et al., reported increased IOP in about 1/3 of vitrectomy performed in eyes that received trabeculectomy, which was corroborated by our study [4]. In our study, IOP was increased after the surgery, as compared to mean IOP before 23G PPV, but the difference not statistically significant. In the case of gas tamponade, the early postoperative IOP was increased but controlled under topical glaucoma medication within 2 weeks. Topical glaucoma medication was used in 4 eyes before the surgery. The number of topical glaucoma medication was increased between preoperative and final follow-up in 2 eyes. Among them, IOP was not adjusted by topical glaucoma medication in 1 eye (proliferative diabetic retinopathy patient) that required additional glaucoma surgery, after which IOP was stabilized. Kunikata et al. reported that the IOP increased to ≥ 20 mmHg in 30 % of subjects after 25G PPV, of which 3 eyes required additional glaucoma surgery. All cases were neovascular glaucoma patients caused by proliferative diabetic retinopathy [7]. Conjunctival scarring occurs in most cases after 20G conventional PPV, but transconjunctival sutureless microincisional PPV minimizes conjunctival scarring to increase the success rate of surgery without special surgical difficulty in additional trabeculectomy. While Thompson et al. reported increased IOP in 1/3 subjects after 20G PPV and occurrence of hypotony in 1/3, we found no case of hypotony with IOP of ≤ 5 mmHg after 23G PPV, similar to the results of Kunikata et al. [4, 7]. The inter-study differences probably resulted from relatively more hypotonic eyes with persistent retinal detachments after vitrectomy reported by Thompson et al. [4].

Thompson et al. reported that bleb function was only maintained in 1/3 of cases where vitrectomy was performed after trabeculectomy [4]. In trabeculectomy, the morphology of bleb was closely related to the success of the surgery. However, there are few reports on researches of the morphological changes of bleb after PPV. Various classification methods were introduced to evaluate the morphology of bleb, and representative methods include classification method by Kronfeld, Indiana bleb appearance grading scale (IBAGS), and Moorfields bleb grading system scale (MBGS) [12–14]. However, such classification methods are based on subjective opinions obtained from slit lamp examination that are not quantitative and cannot reflect the internal structure of the bleb. Ultrasound biomicroscopy (UBM) was the first method introduced to observe internal morphology of the bleb [15]. However, UBM is limited in that it requires contact and axial resolution can reach only 25 μm [16]. On the contrary, anterior segment OCT widely used for clinical purpose is a non-contact and non-invasive method with high resolution of 10 μm. It is the most accurate and easy method for estimating internal structure and function of bleb [17]. Therefore, to evaluate bleb in this study, the hyporeflective space on anterior segment OCT was defined as subconjunctival fluid to compare height. Existence of subconjunctival fluid in the internal structure image indirectly suggests continuous leakage of aqueous humour, and height of fluid in the subconjunctival space can be used as an important factor of bleb function [18]. The height of subconjunctival fluid compared between preoperative, 1 week, 1 month, and 6 months after the surgery in this study showed no statistically significant differences. Transient loss of filtering bleb occurred in 2 eyes (18 %) postoperatively, but it was reformed in 2 ~ 4 weeks. Loss of bleb function may occur temporarily in the early postoperative period. Extensive bleb extending 270° was found in 1 case, and the cannula was inserted through the bleb and conjunctival suture was performed to prevent bleb leakage. There was no difference in IOP before and after the surgery. Conjunctival edema that occurs during surgery negatively affects bleb function by invading the bleb margin; however, we found no differences in IOP before and after the surgery. Therefore, when 23G PPV is done in the trabeculectomized eye, it is helpful to use a pressure plate and adhere space between the conjunctiva and sclera to prevent conjunctival edema. Since there was no significant difference in IOP caused by conjunctival edema, as indicated by the study results, the leaking status was checked to perform conjunctival suture when necessary.

Conventional 20G PPV can be accompanied by injuries such as episcleral and conjunctival vessel, thus affecting filtering bleb function. Hemorrhage can affect the pre-existing filtering bleb. Based on this principle, many existing studies reported autologous blood injection for treatment of overfiltering or leaking blebs after glaucoma surgery [19, 20]. Thus in case of sutureless 23G PPV, subconjunctival hemorrhage is minimized during the surgery to minimize the effect on bleb. Subconjunctival hemorrhage occurred in 27 % of the study subjects, which was similar to the occurrence of subconjunctival hemorrhage in 30 % of cases observed by Kunikata et al. [7]. Care is required to prevent damage to conjunctival vessel during 23G PPV, especially during sclerotomy in eyes that received trabeculectomy. When bleeding occurs at the sclerotomy site, bleeding must be minimized by positioning of epinephrine gauze. Gotzardis reported that diathermy induces temporary adhesion between conjunctiva and sclera, which could reduce the risk of subconjunctival hemorrhage and conjunctival chemosis [21].

The current study had some limitations. First, retrospective analysis commonly has more sources of error due to confounding factors and bias. Second, the sample size was too small because it is difficult to collect a large number of such cases. Better results can be expected by gathering more cases and comparing with the group that received 23G PPV in eyes without history of glaucoma surgery. In addition, we did not compare 23G PPV and 20G PPV.