Abstract
Purpose:
To investigate the isolated organisms, visual acuity (VA) and intraocular pressure (IOP) outcomes associated with delayed-onset bleb-related endophthalmitis (BRE) in eyes treated with pars plana vitrectomy (PPV) versus vitreous biopsy and injection of intravitreal antibiotics (TI).
Methods:
A retrospective chart review was conducted at the Bascom Palmer Eye Institute (Miami, Florida) and University of Florida (Gainesville, Florida) identifying patients treated for BRE with at least one month of follow-up from the inception of an electronic health record (2014 and 2011, respectively) through 2021. All patients had undergone bleb-forming glaucoma surgery at least 1 month prior to endophthalmitis diagnosis
Results:
Thirty-nine eyes from 39 patients (33 from Bascom Palmer, 6 from University of Florida) were included. Trabeculectomy was performed in 34 of 39 eyes (87.2%). Streptococcus species (9 eyes, 23.1%) and Staphylococcus species (8 eyes, 20.5%) were the most common isolated organisms and were similar in both treatment groups (p=0.49). Baseline VA was worse in the PPV group (logMAR 2.51 versus 2.16, p=0.04), but VA at last follow up was similar (p=0.48) in both groups. After recovery from BRE, the average intraocular pressure in the PPV group was 15.1 mmHg on 0.9 IOP-lowering medications compared to 12.6 mmHg on 1.2 medications in the TI group (IOP: P=0.56; medications: P=0.80). Additional glaucoma surgery was performed in 44.4% of the PPV eyes and 16.7% of the TI eyes (P=0.09).
Conclusions:
In eyes with delayed-onset BRE, isolated organisms, visual outcomes and IOP control were similar after initial PPV compared to TI.
Keywords: bleb-related endophthalmitis, trabeculectomy, pars plana vitrectomy, vitreous biopsy
Précis:
Bleb-related endophthalmitis is a serious complication of glaucoma filtration surgery. The current study reports similar visual and intraocular pressure outcomes in eyes treated with pars plana vitrectomy compared to vitreous biopsy and intravitreal antibiotics.
Introduction
Trabeculectomy remains the most commonly performed glaucoma filtration surgery in the United States despite a significant decline among Medicare beneficiaries from 2008 to 2016.1 When successful, trabeculectomy provides excellent intraocular pressure (IOP) control, but also carries a lifelong risk of bleb-related endophthalmitis (BRE), particularly when combined with antifibrotic agents such as mitomycin C or 5-fluorouracil, or when a bleb leak is present.2–4 Rates of endophthalmitis after trabeculectomy are reported as high as 5% after twenty years,5 and BRE can be visually devastating.6,7 Although trabeculectomy has been the reference standard for glaucoma filtration surgery since its introduction in 1968, newer bleb forming procedures have been introduced and can also result in BRE. These include the Ex-PRESS mini-shunt (Alcon, Fort Worth, TX), XEN gel stent (Allergan, Irvine, CA) and PreserFlo Microshunt.8–10 As these procedures increase in utilization, we speculate that the risk of delayed-onset bleb leakage and endophthalmitis is likely to rise.
Treatment for endophthalmitis includes prompt vitreous biopsy and administration of intravitreal antibiotics with or without pars plana vitrectomy (PPV). In patients with acute-onset endophthalmitis following cataract surgery, the Endophthalmitis Vitrectomy Study demonstrated a visual outcome benefit in eyes treated with PPV compared to vitreous biopsy and intravitreal antibiotic injection (TI) in eyes with a visual acuity of light perception or worse.11 In eyes with BRE, however, limited data exists to help guide care. Although a lack of consensus exists, a prior study by Song et al. found that patients with worse visual acuity were more likely to receive PPV as compared to TI.12 The purpose of this study was to investigate the isolated organisms, visual acuity and intraocular pressure outcomes associated with delayed-onset BRE in eyes treated initially with PPV versus TI.
Patients and Methods
A retrospective chart review was conducted to identify all patients treated for BRE at two large tertiary care academic institutions: Bascom Palmer Eye Institute (Miami, Florida) and the University of Florida (Gainesville, Florida). This study was approved by the both the Institutional Review Boards for Human Research at the Universities of Miami and Florida and was conducted in agreement with the provisions of the Declaration of Helsinki. The search was initiated from the inception of electronic medical record implementation at Bascom Palmer Eye Institute and the University of Florida (2014 and 2011, respectively) through December 31, 2021. Patients were included if they had had previously undergone a bleb-forming glaucoma filtering surgery (including trabeculectomy, Ex-PRESS mini-shunt placement, or XEN gel stent implantation) prior to their endophthalmitis diagnosis. Exclusion criteria consisted of eyes with acute-onset endophthalmitis that developed within one month of glaucoma surgery, and eyes with less than one month follow-up after treatment of endophthalmitis. Eyes that required evisceration at any time point were not excluded.
A diagnosis of bleb-associated endophthalmitis was made based upon the clinical features of diffuse bulbar conjunctival hyperemia, fibrino-purulent inflammatory precipitate within the filtering bleb, and marked intraocular inflammation characterized by hypopyon and vitritis. Vitritis was established based upon slit lamp and dilated funduscopic examination, and confirmed with B-scan ultrasonography demonstrating inflammatory precipitates in the vitreous body. Patients with minimal anterior chamber reaction and without vitritis were not considered to have endophthalmitis. All patients underwent either primary PPV or TI at the discretion of the treating surgeon. Bacteria were identified with a combination of Gram stain and vitreous culture. Vitreous fluid was inoculated directly onto fresh aerobic and anaerobic blood agar, chocolate agar, Sabouraud agar and thioglycolate broth. Vitrectomy specimens were filtered through a 0.45 micron filter, sectioned and placed on media as above. Alternatively 10–20 ml. of vitrectomy specimen was inoculated into blood culture bottles. All media except the Sabouraud agar were incubated at 37 degrees Celsius. Sabouraud agar was incubated at room temperature (25 degrees Celsius). Drops were placed on slides and processed with Gram’s and Giemsa stains; microbiology logs were examined to confirm final culture results. All eyes received intravitreal and topical antibiotics. Intravitreal and topical corticosteroids were variably utilized.
Medical records were reviewed to identify patient age, gender, race and ethnicity, type and location of glaucoma filtering surgery, eye, preoperative diagnosis, follow-up period, preoperative and postoperative medications, preoperative and postoperative IOP, and postoperative course. Clinical outcomes were recorded with special attention to additional interventions required to eradicate infection, and additional surgery to maintain IOP control. Eyes that required more than one treatment, such as a second TI, conversion from TI to PPV, or underwent evisceration for persistent infection were classified as having an inadequate initial treatment.
Statistical Analysis
Visual acuity of count fingers, hand motion, light perception, and no light perception were assigned to logMAR equivalents of 2.0, 2.3, 2.6, and 3.0, respectively for statistical analysis. Descriptive statistics for categorical variables include counts and percentages and for continuous variables include mean, standard deviation, and range. Treatment group differences were assessed with Fisher exact or exact chi-square tests (for categorical variables); the Jonckheere-Terpstra test (for ordinal category variables), and the Mann-Whitney Wilcoxon (MWW) two-sample test (for continuous, non-normally distributed variables). Final VA differences by culture status were also assessed with the MWW test. All analyses were done using SAS version 9.4 software (SAS Institute, Cary, NC, USA). A p-value ≤ 0.05 was considered statistically significant.
Results
The clinical characteristics of the study population are presented in Table 1. A total of 39 eyes from 39 patients (33 from Bascom Palmer Eye Institute, 6 from the University of Florida) were selected for review. Figure 1 illustrates a slit lamp photograph corresponding to case 16 illustrates an ischemic and purulent bleb with diffuse conjunctival injection and hypopyon at presentation with endophthalmitis (A), and three weeks after vitreous biopsy and treatment with intravitreal antibiotics (B). This patient had significant vitritis on B-scan ultrasonography with dense vitreous membranes and a tractional retinal detachment, which contributed to the poor visual outcome. Nine eyes (23.1%) underwent PPV, and thirty eyes (76.9%) underwent TI. Table 2 summarizes the patient characteristics in eyes treated with pars plana vitrectomy compared to vitreous tap and injection of intravitreal antibiotics. No significant differences were observed between the two treatment groups. Among those eyes that underwent TI, nine eyes (30.0%) had subsequent PPV including seven eyes with persistent infection, and two eyes with concurrent retinal detachment.
Table 1:
Clinical data in patients with bleb-related endophthalmitis
| Patient Number | Time Since Glaucoma Surgery (months) | Type of Glaucoma Surgery | Causative Organism | Initial Treatment | Baseline VA (logMAR) | Last VA (logMAR) | Follow-up (months) | Last IOP | Glaucoma Medications at Last Follow-Up (N) | Additional Glaucoma Surgery Required | Re-treatment of Endophthalmitis | Phthisis (IOP <5 mm Hg) | Evisceration/Enucleation |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 240 | Trabeculectomy | Streptococcus gordonii | PPV | 2.3 | 2.0 | 25 | 13 | 0 | No | No | No | No |
| 2 | N/A | Trabeculectomy | Methacillin-resistant Staphylococcus aureus | PPV | 2.3 | 2.3 | 42 | 20 | 1 | GDD | No | No | No |
| 3 | 2 | Ex-PRESS Trabeculectomy | Coagulase-negative Staphylococcus | TI | 0.4 | 0.5 | 1 | 14 | 0 | No | No | No | No |
| 4 | 144 | Trabeculectomy | Culture negative | PPV | 2.3 | 1.6 | 1 | 4 | 0 | No | No | No | No |
| 5 | 36 | Ex-PRESS Trabeculectomy | Culture negative | PPV | 2.6 | 2.0 | 35 | 11 | 4 | CPC | No | No | No |
| 6 | N/A | Trabeculectomy | Streptococcus gordonii | PPV | 2.6 | 1.176 | 81 | 12 | 1 | Bleb revision | No | No | No |
| 7 | 84 | Trabeculectomy | Culture negative | PPV | 2.3 | 3.0 | 53 | 26 | 1 | No | No | No | No |
| 8 | 25 | Trabeculectomy | Culture negative | PPV | 2.6 | 3.0 | 41 | 29 | 0 | No | No | No | No |
| 9 | 12 | Trabeculectomy | Culture negative | PPV | 2.6 | 2.3 | 54 | 6 | 0 | GDD | No | No | No |
| 10 | N/A | Trabeculectomy | Enterococcus faecalis | PPV | 3.0 | 3.0 | 1 | N/A | N/A | No | Yes | No | Evisceration |
| 11 | 60 | Trabeculectomy | Culture negative | TI | 2.3 | 2.3 | 52 | 13 | 2 | No | Yes | No | No |
| 12 | 60 | Trabeculectomy | Culture negative | TI | 2.6 | 3.0 | 0.25 | N/A | N/A | No | Yes | No | Enucleation |
| 13 | 20 | XEN Gel Stent | Culture negative | TI | 2.3 | 3.0 | 3 | 16 | 0 | No | Yes | No | No |
| 14 | N/A | Trabeculectomy | Culture negative | TI | 3.0 | 3.0 | 13 | N/A | N/A | No | Yes | No | Evisceration |
| 15 | 153 | Trabeculectomy | Culture negative | TI | 1 | 0.3 | 30 | 15 | 4 | No | No | No | No |
| 16 | 156 | Trabeculectomy | Culture negative | TI | 1.3 | 2.3 | 52 | 12 | 5 | Bleb revision | Yes | No | No |
| 17 | 240 | Trabeculectomy | Culture negative | TI | 2.3 | 3.0 | 3 | 36 | 0 | No | Yes | No | No |
| 18 | 120 | Trabeculectomy | Culture negative | TI | 2.6 | 2.6 | 1 | 0 | 1 | No | No | No | No |
| 19 | 120 | Trabeculectomy | Staphylococcus epidermidis | TI | 2.3 | 1 | 4 | 14 | 0 | No | Yes | No | No |
| 20 | 72 | Trabeculectomy | Streptococcus viridans | TI | 2.3 | 0.7 | 67 | 13 | 3 | Bleb revision with GDD | No | No | No |
| 21 | 120 | Trabeculectomy | Streptococcus pneumoniae | TI | 2.6 | 3.0 | 45 | N/A | N/A | No | No | No | Evisceration |
| 22 | N/A | Trabeculectomy | Streptococcus mitis | TI | 2.3 | 2.6 | 11 | 2 | 0 | No | No | No | No |
| 23 | N/A | Trabeculectomy | Culture negative | TI | 2.3 | 0.177 | 21 | 13 | 3 | No | No | No | No |
| 24 | 84 | Trabeculectomy | Alpha hemolytic Streptococcus | TI | 2.3 | 2.6 | 73 | 18 | 1 | No | Yes | No | No |
| 25 | 60 | Trabeculectomy | Streptococcus pneumoniae | TI | 2.3 | 3.0 | 51 | 0 | 0 | No | No | Yes | No |
| 26 | 15 | Trabeculectomy | Streptococcus mitis | TI | 2.3 | 2.3 | 52 | 4 | 0 | Bleb revision | No | No | No |
| 27 | 1 | Phacoemulsification and trabeculectomy | Streptococcus mitis | TI | 2.6 | 3.0 | 50 | 42 | 0 | No | Yes | No | No |
| 28 | 1 | Trabeculectomy | Staphylococcus epidermidis | TI | 2.3 | 2.3 | 1 | 16 | 3 | No | No | No | No |
| 29 | 120 | Trabeculectomy | Staphylococcus aureus | TI | 2.3 | 1 | 2 | 13 | 1 | No | No | No | No |
| 30 | 360 | Trabeculectomy | Staphylococcus epidermidis | TI | 2.3 | 0.3 | 11 | 11 | 3 | No | Yes | No | No |
| 31 | 36 | Trabeculectomy | Culture negative | TI | 2.3 | 2.0 | 62 | 10 | 0 | No | No | No | No |
| 32 | 84 | Trabeculectomy | Staphylococcus aureus | TI | 2.0 | 0.3 | 34 | 9 | 0 | No | No | No | No |
| 33 | 168 | Trabeculectomy | Culture negative | TI | 0.4 | 0.4 | 51 | 9 | 0 | No | No | No | No |
| 34 | N/A | Trabeculectomy | Serratia marcescens | TI | 2.0 | 2.3 | 42 | 20 | 1 | Bleb revision with GDD | No | No | No |
| 35 | 60 | Trabeculectomy | Culture negative | TI | 2.3 | 1.6 | 1 | 6 | 0 | No | No | No | No |
| 36 | 73 | Trabeculectomy | Culture negative | TI | 3.0 | 3.0 | 13 | 0 | 0 | No | No | Yes | No |
| 37 | N/A | Trabeculectomy | Culture negative | TI | 2.3 | 2.3 | 9 | 10 | 3 | No | No | No | No |
| 38 | 13 | Ex-PRESS Trabeculectomy | Staphylococcus aureus | TI | 2.3 | 1.3 | 2 | 14 | 0 | Bleb revision | No | No | No |
| 39 | 300 | Trabeculectomy | Gram variable rods | TI | 2.3 | 0.3 | 63 | 10 | 2 | No | No | No | No |
PPV = pars plana vitrectomy, TI = vitreous tap and injection of intravitreal antibiotics, VA = visual acuity, IOP = intraocular pressure, GDD = glaucoma drainage device, CPC = cyclophotocoagulation
Figure 1:
Slit lamp photograph corresponding to case 16 illustrates an ischemic and purulent bleb with diffuse conjunctival injection and hypopyon at presentation with endophthalmitis (A), and three weeks after vitreous biopsy and treatment with intravitreal antibiotics (B).
Table 2:
Patient characteristics in eyes treated with pars plana vitrectomy compared to vitreous tap and injection of intravitreal antibiotics
| All Eyes (n=39) | PPV (n=9) | TI (n=30) | P-value | |
|---|---|---|---|---|
| Mean age (years) | 72.5 (SD 13.4 range 37–92) | 78.0 (SD 9.4, range 64–92) | 71.3 (SD 14.1 range 37–87) | 0.13 |
| Ocular laterality | 0.12 | |||
| Right | 21 (53.8%) | 3 (33.3%) | 18 (60.0%) | |
| Left | 18 (46.2%) | 6 (66.7%) | 12 (40%) | |
| Gender | 0.24 | |||
| Male | 21 (53.8%) | 4 (44.4%) | 17 (56.7%) | |
| Female | 18 (46.2%) | 5 (55.6%) | 13 (43.3%) | |
| Glaucoma diagnosis | 0.48 | |||
| POAG | 27 (69.2%) | 7 (77.8%) | 20 (66.7%) | |
| JOAG | 1 (2.6%) | 0 | 1 (3.3%) | |
| XFG | 1 (2.6%) | 1 (11.1%) | 0 | |
| Unknown | 10 (25.6%) | 1 (11.1%) | 9 (30.0%) | |
| Glaucoma surgery | 1.00 | |||
| Trabeculectomy | 35 (89.7%) | 8 (88.9%) | 27 (90.0%) | |
| Ex-PRESS shunt | 3 (7.7%) | 1 (11.1%) | 2 (6.7%) | |
| XEN gel stent | 1 (2.6%) | 0 | 1 (3.3%) | |
| Mean time interval between glaucoma surgery and endophthalmitis (months) | 98.0 (SD 89.7, range 1–360) | 90.2 (SD 88 range 12 –240) | 99.1 (SD 91.8 range 1–360) | 0.86 |
| Bleb leak present | 9 (23.1%) | 1 (11.1%) | 8 (26.7%) | 0.71 |
| Mean follow-up (months) | 29.6 (SD 25.1 range 0.25 –81) | 37.0 (SD 25.7) range 1–81 | 27.9 (SD 24.9 range 0.25–67) | 0.45 |
PPV = pars plana vitrectomy, TI = vitreous biopsy with intravitreal antibiotic injection, SD = standard deviation, POAG = primary open-angle glaucoma, JOAG = juvenile open angle glaucoma, XFG = exfoliation glaucoma
The underlying isolated organisms for BRE are summarized in Figure 2. The most common organisms were Streptococcus species (9 eyes, 23.1%) and Staphylococcus species (8 eyes, 20.5%), and were similar in both treatment groups (p=0.5). Table 3 demonstrates the clinical outcomes based upon the culture results in eyes with BRE. No eyes that presented with a VA of light perception or worse were due to Staphylococcal disease. A higher percentage of eyes with a VA at last follow up of count fingers or worse were due to Streptococcal versus Staphylococcal disease (28.0% versus 8.0%). Visual acuity at last follow-up in culture negative eyes (n=20) compared to culture positive eyes (n=19) were similar (mean logMAR 2.06 ± 1.02, mean logMAR 1.83 ± 0.98, p=0.51). There were no trends between underlying isolated organism and the type of treatment initially pursued (p=0.5).
Figure 2:
Pie chart demonstrating the isolated organisms (%) of bleb-related endophthalmitis amongst the study population (n=39).
Table 3:
Clinical outcomes based upon culture results in eyes with bleb-related endophthalmitis
| Outcome | Culture Result |
|---|---|
| All patients | 4 (10.3%) Staphylococcus aureus 3 (7.7%) Streptococcus mitis 3 (7.7%) Staphylococcus epidermidis 2 (5.1%) Streptococcus pneumoniae 2 (5.1%) Streptococcus gordonii 1 (2.6%) Coagulase-negative Staphylococcus 1 (2.6%) Enterococcus faecalis 1 (2.6%) Streptococcus viridans 1 (2.6%) Serratia marascens 1 (2.6%) Alpha-hemolytic Streptococcus 20 (51.3%) Culture negative |
| Presenting visual acuity hand motion or better | 8 (28.6%) Staphylococcus species 6 (21.4%) Streptococcus species 1 (3.6%) Serratia marascens 13 (46.4%) Culture negative |
| Presenting visual acuity light perception or worse | 3 (27.3%) Streptococcus species 1 (9.1%) Enterococcus faecalis 7 (63.6%) Culture negative |
| Final visual acuity 20/400 or better | 6 (42.9%) Staphylococcus species 2 (14.3%) Streptococcus species 6 (42.9%) Culture negative |
| Final visual acuity count fingers or worse | 7 (28.0%) Streptococcus species 2 (8.0%) Staphylococcus species 1 (4.0%) Serratia marascens 1 (4.0%) Enterococcus faecalis 14 (56.0%) Culture negative |
| Eyes treated with pars plana vitrectomy | 2 (22.2%) Streptococcus species 1 (11.1%) Enterococcus faecalis 1 (11.1%) Staphylococcus species 5 (55.6%) Culture negative |
| Eyes treated with vitreous tap and intravitreal antibiotics | 7 (23.3%) Staphylococcus species 7 (23.3%) Streptococcus species 1 (3.3%) Serratia marascens 15 (50.0%) Culture negative |
| Eyes requiring re-treatment for endophthalmitis | 2 (18.2%) Streptococcus species 2 (18.2%) Staphylococcus species 1 (9.1%) Enterococcus faecalis 6 (54.5%) Culture negative |
The visual and IOP outcomes for all patients, as well as separately for patients who underwent an initial PPV versus a TI, are presented in Table 4. Eyes that underwent PPV had significantly (p=0.04) worse baseline vision compared to eyes that underwent TI. All other visual and IOP outcome data were similar (p>0.05) between the two groups. Among patients who underwent additional glaucoma surgery, four underwent bleb revision, two had glaucoma drainage device (GDD) implantation surgery, two underwent bleb revision and GDD surgery, and one had cyclophotocoagulation.
Table 4:
Comparison of outcomes in eyes with bleb-related endophthalmitis treated with pars plana vitrectomy compared to vitreous tap and injection of intravitreal antibiotics
| Outcome | All Patients (n=39) | PPV (n=9) | TI (n=30) | P-value |
|---|---|---|---|---|
| Evisceration | 4 (10.3%) | 1 (11.1%) | 3 (10.0%) | 0.44 |
| Phthisis | 2 (5.1%) | None | 2 (6.7%) | 0.59 |
| Need for additional glaucoma surgery | 9 (23.1%) | 4 (44.4%) | 5 (16.7%) | 0.09 |
| Need for additional endophthalmitis treatment | 11 (28.2%) | 1 (11.1%) | 10 (33.3%) | 0.16 |
| Mean (SD) LogMAR visual acuity at time of endophthalmitis diagnosis | 2.24 (SD 0.57, range 0.4–3.0 ) | 2.51 (SD 0.24, range 2.0–3.0) | 2.16 (SD 0.61, range 0.4–3.0) | 0.04 |
| Mean (SD) LogMAR final visual acuity | 1.94 (SD 0.99, range 0.18–3.0) | 2.26 (SD 0.65, range 1.2–3.0) | 1.86 (SD 1.07, range 0.18–3.0) | 0.48 |
| Mean (SD) LogMAR change in visual acuity | −0.3 (SD 0.84) | −0.25 (SD 0.63) | −0.31 (SD 0.9) | 0.80 |
| Mean (SD) IOP at final follow-up (mmHg) | 13.2 (SD 9.2, range 0–42) | 15.1 (SD 9.0, range 6–29) | 12.6 (SD 9.4, range 0–42) | 0.56 |
| Mean (SD) number of glaucoma medications at final follow-up | 1.11 (SD 1.47, range 0–5) | 0.88 (SD 1.36, range 0–4) | 1.14 (SD 1.52 range 0–5) | 0.80 |
PPV = pars plana vitrectomy, TI = vitreous tap and injection of intravitreal antibiotics, SD = standard deviation
Discussion
BRE represents one of the most visually devastating complication of glaucoma filtration surgery. Fifty percent of cases initially present as localized blebitis, as the fistula provides ready access from the bleb into the anterior chamber and thus the potential for rapid dissemination of infection.13 Because of the rapidity with which the infection can progress, guidelines were suggested by Collaborative Bleb-Related Infection and Treatment Study14 which recommended urgent antibiotic therapy in eyes with BRE and mild vitreous opacities, whereas those with more marked vitreous involvement were recommended to have a PPV. There is a lack of published data to support these guidelines, and new glaucoma surgical procedures have emerged which generate conjunctival filtering blebs raising the risk for future infections. The present study adds a contemporary series (2011–2021) of patients undergoing treatment for delayed onset BRE and provides current data regarding the visual and intraocular pressure outcomes in eyes with BRE treated with TI versus PPV.
In the current study, the most common organisms responsible for BRE were Streptococcus and Staphylococcus species. This is similar to prior studies, including Leng et al.’s cohort of 71 eyes, in which 30% of cases were due to Streptococcus species, 28% due to gram-negative organisms, and 18% due to Staphylococcus species.6 Song et al. reported the most common organisms to be Streptococcus (31%) and Staphylococcus (22%) species in their series of 49 eyes.12 In the current series, eyes with Streptococcus and Enterococcus species were more likely to have worse visual outcomes and present with LP vision. Staphylococcus species were associated with a VA at last follow up of 20/400 or better.
Overall, clinical outcomes were similar between the PPV and TI groups with respect to visual acuity and IOP control. There was a trend towards higher rates of requiring additional surgery for IOP control in the PPV group, but this was not statistically significant. There was also a trend towards more successful eradication of infection on the first intervention with PPV than TI. Over one-quarter of patients who underwent an initial TI ultimately underwent a PPV in the current study. This is consistent with the rate reported by Leng et al6, in which 40% of patients who underwent a primary TI ultimately required a PPV, as well as the 30% rate reported by Song et al.12 Visual acuity outcomes were relatively poor in both groups, with most patients having a VA at last follow up of count fingers or worse. Of note, the VA at endophthalmitis diagnosis was worse in the PPV group, and yet this group achieved similar VA outcomes to the TI group. This supports the concept that PPV is currently used for more virulent infections and also that it may be better at preserving VA than TI. A prospective, randomized clinical trial, perhaps preceded by a retrospective analysis of Intelligent Research In Sight (IRIS) Registry® data, would be necessary to support this theory.
The current study has limitations that include a retrospective study design, as well as a relatively small sample of eyes that received initial PPV. As patients were not randomized to treatment, the potential for selection bias exists with preferences towards treatment with primary PPV versus TI. However, BRE is an uncommon complication of glaucoma filtering surgery which makes a prospective randomized trial challenging to conduct. Risk factor recognition and treatment for late-onset bleb leakage, and reductions in both the concentration and duration of antifibrosis therapy application during glaucoma filtering surgery, have contributed to this reduced incidence. Thus, the current study provides important information regarding clinical outcomes with both treatment approaches.
In conclusion, BRE is a serious complication of glaucoma filtration surgery. The current study reports similar isolated organisms, visual outcomes and IOP control in eyes treated with pars plana vitrectomy compared to vitreous biopsy and intravitreal antibiotics.
Funding Sources:
P30EY014801 University of Miami core grant; Research to Prevent Blindness Unrestricted grant, New York, New York; an unrestricted grant from Mr. and Mrs. Thorne B. Donnelley, Chicago, IL, the Maltz Family Endowment for Glaucoma Research, Cleveland, Ohio. Research reported in this publication was also supported by the University of Florida Clinical and Translational Science Institute, which is supported in part by the NIH National Center for Advancing Translational Sciences under award number UL1TR001427. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
Publisher's Disclaimer: Disclosures: The authors declare that they have no conflicts of interest.
References
- 1.Rathi S, Andrews CA, Greenfield DS, Stein JD. Trends in glaucoma surgeries performed by glaucoma subspecialists versus nonsubspecialists on Medicare beneficiaries from 2008 through 2016. Ophthalmology. 2021;128(1):30–38. doi: 10.1016/j.ophtha.2020.06.051 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Higginbotham EJ, Stevens RK, Musch DC, et al. Bleb-related endophthalmitis after trabeculectomy with mitomycin C. Ophthalmology. 1996;103(4):650–656. doi: 10.1016/S0161-6420(96)30639-8 [DOI] [PubMed] [Google Scholar]
- 3.Greenfield DS. Endophthalmitis after filtering surgery with mitomycin. Arch Ophthalmol. 1996;114(8):943. doi: 10.1001/archopht.1996.01100140151007 [DOI] [PubMed] [Google Scholar]
- 4.Lama PJ, Fechtner RD. Antifibrotics and wound healing in glaucoma surgery. Survey of Ophthalmology. 2003;48(3):314–346. doi: 10.1016/S0039-6257(03)00038-9 [DOI] [PubMed] [Google Scholar]
- 5.Olayanju JA, Hassan MB, Hodge DO, Khanna CL. Trabeculectomy-related complications in Olmsted county, Minnesota, 1985 through 2010. JAMA Ophthalmol. 2015;133(5):574. doi: 10.1001/jamaophthalmol.2015.57 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Leng T, Miller D, Flynn HW, Jacobs DJ, Gedde SJ. Delayed-onset bleb-associated endophthalmitis (1996–2008): Causative organisms and visual acuity outcomes. Retina. 2011;31(2):344–352. doi: 10.1097/IAE.0b013e3181e09810 [DOI] [PubMed] [Google Scholar]
- 7.Ohtomo K, Mayama C, Ueta T, Nagahara M. Outcomes of late-onset bleb-related endophthalmitis treated with pars plana vitrectomy. Journal of Ophthalmology. 2015;2015:1–9. doi: 10.1155/2015/923857 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Stewart RM, Diamond JG, Ashmore ED, Ayyala RS. Complications Following Ex-Press glaucoma shunt implantation. American Journal of Ophthalmology. 2005;140(2):340.e1–340.e3. doi: 10.1016/j.ajo.2005.02.033 [DOI] [PubMed] [Google Scholar]
- 9.Olgun A, Imamoğlu S, Karapapak M, Düzgün E, Kaçar H. Endophthalmitis after XEN gel stent implantation: 2 cases. Journal of Glaucoma. 2018;27(12):e191–e194. doi: 10.1097/IJG.0000000000001076 [DOI] [PubMed] [Google Scholar]
- 10.Brambati M, Bettin P, Ramoni A, Battista M, Bandello F. A case of endophthalmitis following needling procedure after PRESERFLO ® Micro Shunt implantation. European Journal of Ophthalmology. Published online May 20, 2021:112067212110195. doi: 10.1177/11206721211019548 [DOI] [PubMed] [Google Scholar]
- 11.Results of the Endophthalmitis Vitrectomy Study. A randomized trial of immediate vitrectomy and of intravenous antibiotics for the treatment of postoperative bacterial endophthalmitis. Endophthalmitis Vitrectomy Study Group. Arch Ophthalmol. 1995;113(12):1479–1496. [PubMed] [Google Scholar]
- 12.Song A, Scott IU, Flynn HW, Budenz DL. Delayed-onset bleb-associated endophthalmitis: Clinical features and visual acuity outcomes. Ophthalmology. 2002;109(5):985–991. doi: 10.1016/s0161-6420(02)00965-x [DOI] [PubMed] [Google Scholar]
- 13.Yamamoto T. Bleb-related infection: Clinical features and management. Taiwan Journal of Ophthalmology. 2012;2(1):2–5. doi: 10.1016/j.tjo.2011.12.003 [DOI] [Google Scholar]
- 14.Yamamoto T, Kuwayama Y. Interim clinical outcomes in the Collaborative Bleb-Related Infection Incidence and Treatment Study. Ophthalmology. 2011;118(3):453–458. doi: 10.1016/j.ophtha.2010.07.002 [DOI] [PubMed] [Google Scholar]