Infectious endophthalmitis is a serious vision-threatening condition. This course reviews current trends in etiology, causative microorganisms, management and evolving surgical techniques to address this condition.
Infectious endophthalmitis is a serious vision-threatening condition involving inflammation of the globe and its contents. It most commonly occurs as a post-surgical or post-traumatic complication; however, it can also occur via an endogenous source through the bloodstream. Endophthalmitis requires early intervention to avoid potentially irreversible vision loss. Its diagnosis is based primarily on clinical exam findings. Prompt administration of intravitreal (IVT) antibiotics, with pars plana vitrectomy (PPV) in certain circumstances, has been considered standard management ever since the results of the Endophthalmitis Vitrectomy Study (EVS) were published in 1995. Here we will review current trends in etiology, causative microorganisms, management and evolving surgical techniques for post-operative infectious exogenous endophthalmitis, with special focus on cataract surgery, IVT injections, and glaucoma surgery.
Infectious endophthalmitis can be categorized into several subtypes. Large retrospective studies have identified the most common types as post-surgical (31.2%), bleb-associated (15.6%), corneal ulcer-related (13.6%), endogenous (13.1%) and post-IVT injection (most commonly related to vascular endothelial growth factor VEGF inhibitors; 8.5%).1 Here we will surmise the several exogenous subtypes.
The most common form of endophthalmitis is acute infectious endophthalmitis (AIE) following cataract surgery. Endophthalmitis presenting within six weeks of cataract surgery is considered to be acute. Given that cataract surgery is one of the most commonly performed surgeries worldwide, this is not surprising. Prior to modern cataract extraction techniques, it was seen in as many as 2% of surgical cases.2 With the advent of antibiotics, aseptic techniques and newer surgical approaches, reported rates dropped to 0.086% to 0.110%.2 More recent studies have found even lower rates in countries that began using intracameral (IC) cefuroxime at the end of surgery.3 Since the year 2000, rates as low as 0.012% have been reported.4
Recognized risk factors for the development of endophthalmitis following cataract surgery include: older age, male gender, extra- or intracapsular cataract extraction versus phacoemulsification, clear corneal incision, failing to use IC cefazolin or cefuroxime, posterior capsular rupture (PCR), silicone intraocular lens (IOL) and intraoperative complications (e.g., posterior capsule rupture, vitreous loss).4 Other patient-related risk factors include immunosuppression, blepharitis, and diabetes mellitus.2,3,5
In the original EVS publication, Gram-positive cocci were identified as the most common infective cause for endophthalmitis following cataract surgery,6 with Staphylococcus epidermidis being the most common species.6,7 More broadly, Gram-positive organisms have been reported to account for 80.5% of culture-positive isolates,8 with Gram-negatives accounting for only 7.1 to 11.0%.8,9 When broken down further within Gram-positives, Coagulase-negative Staphylococci (CoNS), which include S. epidermidis, comprise 54.6%, while Streptococcus species (20.8%), Staphylococcus aureus (10.2%), and other Gram-positives (7.4%) are also frequently isolated.9-11 Fungi and yeast make up 8.4% of isolates.8
International studies have found similar results and microbiological spectrum in most countries throughout the world, including the United States, Europe, and Australia.12 However, diversity of microbiota can still be found in some areas of the world. Emergence of Enterococcus fecalis as a source of infectious endophthalmitis was seen in South Korea, causing 20.8% of culture-proven cases.13 Regional differences are also well described in India, where there is a similar overall incidence but altered spectrum. Gram-positive cocci only account for 44% to 64.8% of cases (S. aureus most common) and Pseudomonas is increased at 26.2% to 43%.14 In northern India, fungi account for 16.7% to 70% of cases,14 most commonly Aspergillus.14,15
Prompt recognition of infection is critical due to the poor prognosis of infectious endophthalmitis, making time to presentation a theoretical key factor in outcomes. In the original EVS study the median time to presentation was six days, but 22% presented between two and six weeks after cataract surgery.6 Other large studies have found presentation ranging from 8 to 13 days.7,16
Chronic, or delayed-onset, endophthalmitis following intraocular surgery typically presents after six weeks or longer. It often presents as low-grade inflammation in the anterior chamber with mild eye pain.17 Occasionally, hypopyon, lens capsule plaques, and vitreous inflammation may be present.17 Risk factors associated with delayed-onset endophthalmitis include ocular surface disease and retained lens material.2 The most commonly cited causative pathogen is Propionibacterium acnes, with CoNS and fungi as other potential culprits.2,3,17 The less virulent agents tend to lead to a more slowly progressive course. A special consideration during the workup of delayed-onset cases is to perform longer incubation of cultures for up to fifteen days to allow for slower growth organisms to manifest.3
Intravitreal injections have become the most commonly performed procedure in the field of ophthalmology over the past decade, leading to a dramatic increase in the number of post-IVT injection endophthalmitis cases.2,18 Specifically, the use of VEGF antagonists has become a mainstay for the treatment of exudative age-related macular degeneration, diabetic macular edema, diabetic retinopathy, and macular edema associated with retinal vein occlusion.19 In 2016, over 3 million injections were performed in the United States, and this number is only going to increase. Rates and outcomes of infectious endophthalmitis do not appear to vary significantly between the anti-VEGF agents bevacizumab, ranibizumab, and aflibercept.20-22 Although the overall incidence rate is low following IVT injections, commonly reported between 0.016% and 0.056%,19-21,23,24 the volume of injections being delivered has made it one of the most common causes of infectious endophthalmitis.
In cases of endophthalmitis after IVT injection, the source of infection is most often believed to be the patient’s own ocular surface, conjunctiva, and eyelids.6 Recently, evidence suggests patients are at increased risk of oral sources of post-procedure infection.25 The most commonly identified causative agent was CoNS, with Streptococcus species also frequently seen.1,24,26 Streptococcus viridians was specifically associated with more rapid onset and progression.1,24 A 10-year retrospective study of endophthalmitis following anti-VEGF injections found the time between injection and symptoms was 3.8 days on average, and time until presentation after symptom onset was 4.4 days.19 Additionally, the authors found that younger age and lower intraocular pressure (IOP) were predictive of better visual outcomes at 6-months post-treatment.1
Large database studies of endophthalmitis after IVT reveals overall rates of endophthalmitis are consistent with clinical trials. Pooled average rate for all agents is approximately 1 in 2,771 (0.036%). No difference exists for the different VEGF antagonists; however, the highest rate is seen in the steroid medications with the dexamethasone implant (Ozurdex) having an endophthalmitis rate of 0.107% and triamcinolone 0.147%.
Database studies align with other endophthalmitis after IVT literature with a mean diagnosis at 4.7 days from the offending intravitreal injection and this is similar amongst all drugs. Patients lose an average of 74 letters and the mean time to visual recovery within 15 letters is 41 days, and within 5 letters occurring at 66 days. Visual recovery is similar for all of the drugs except triamcinolone, which occurs faster, even though triamcinolone has a purported higher rate of suspected endophthalmitis compared to other intravitreal medications.
Presently, prefilled syringes have increased in popularity owing to efficiency. Nonetheless, debate continues on whether the use of prefilled syringes results in a lower rate of endophthalmitis and this is a current area of growing research.
Trabeculectomy and glaucoma drainage implants (GDI) are common surgical interventions for glaucoma. Endophthalmitis as a complication of glaucoma filtration surgery can occur due to implant exposure, erosion, or bleb access. Bleb-related endophthalmitis (BRE) involves inflammation of the vitreous and is thought to exist on a spectrum of bleb-related infections (BRI).27 For these cases, causative microorganisms are similar between BRE and blebitis.28 Risk factors for endophthalmitis following GDI include implant exposure and recent surgery.29 Additional risk factors for any BRI include pigmentary glaucoma, juvenile glaucoma, bleb leak, sustained low IOP, chronic blepharitis and punctal plugs.30 Mitomycin C and other antimetabolites, which are commonly used during glaucoma surgery to prevent wound scarring and bleb failure, may also lead to increased rates of BRI.31,32
Glaucoma surgery is unique compared to other post-procedural causes of endophthalmitis, which typically present acutely or subacutely, by having an ongoing long-term risk. The cumulative risk is still relatively low, but nevertheless concerning due to the poor visual prognosis. The incidence tends to be greater following trabeculectomy than following GDI.33 Cumulative incidences of BRE over five years is between 0.45% and 1.3% (United States data).34 The Collaborative Initial Glaucoma Treatment Study reported a five-year risk of developing endophthalmitis after trabeculectomy of 1.1%.35 A long-term look at all BRIs following trabeculectomy revealed a rate of less than 2% at a minimum of one year follow-up.30 Following implantation surgery, approximately 0.7% developed endophthalmitis at an average of 2.6 years.36 Large reviews assign post-drainage implant cumulative incidences of 0.22%, 0.43%, and 0.83% at one, five, and ten years, respectively.37
Staphylococcus epidermidis is the most common organism isolated in endophthalmitis following GDI.29 The spectrum may be altered in delayed-onset cases following glaucoma surgery. Whereas CoNS are the most common for early-onset BRE,38,39 Streptococcus species are represented more often than Staphylococcus in delayed cases.38-40 Outcomes are generally poor for BRE.
Despite the rarity of endophthalmitis, its gravity and poor outcomes make avoidance an ongoing area of interest. Prophylaxis has been examined over time and best practices have evolved. Current evidence-based support exists for conjunctiva preparation with 5% povidone-iodine having significantly lower rates of endophthalmitis compared to silver protein solutions.41 Prophylactic topical antibiotics remain controversial and povidone-iodine remains the only technique to receive an intermediate or better clinical recommendation to support its use3,42; topical antibiotics, lash trimming, and saline irrigation have not.3,42 Nevertheless, the 2014 ASCRS survey revealed that 90% of surgeons still use topical antibiotics perioperatively,Chang most commonly a fourth-generation fluoroquinolone.George
More recent debate has centered on the use of IC antibiotics at the time of cataract surgery, with Sharma et al43 finding only a marginal benefit. The original ESCRS study found a 4.92-fold higher rate of endophthalmitis in a group that did not receive IC cefuroxime.3 Schwartz et al44 argue that despite this finding they should not be used as standard of care, in large part due to the high percentage of endophthalmitis cases in the group that did not receive IC antibiotics in the ESCRS study, as well as a lack of other randomized clinical trials (RCTs). Other concerns associated with routine antibiotic use include toxicity related to compounding errors, unnecessary cost and promoting resistance. However, a 2017 systematic review45 of five relevant RCTs investigating efficacy of perioperative antibiotics in cataract surgery concluded that there is high certainty that cefuroxime injection lowers the likelihood of post-surgical endophthalmitis. Multiple studies comparing rates of endophthalmitis from time periods before and after IC cefuroxime implementation have shown significantly reduced rates since its introduction.46,47 A shortcoming of retrospective studies is the failure to account for improvements in surgical and aseptic technique that occurred over the same time periods.George
Recently, both IC moxifloxacin and cefuroxime have demonstrated efficacy at reducing endophthalmitis after phacoemulsification compared to controls.48 A large analysis of cataract surgeries showed that eyes receiving IC moxifloxacin prophylaxis had significantly reduced endophthalmitis rates, three-fold lower in manual small incision cataract surgery (M-SICS) and six-fold lower in phacoemulsification.5 These findings are of clinical interest when considering that moxifloxacin is more commercially available in some countries than cefuroxime. Surveys on European practices following the release of ESCRS recommendations revealed that 54% to 74% of respondents routinely used IC antibiotics,49,50 with a lack of appropriate cefuroxime preparations as the primary reason for not doing so.49 In the United States, cefuroxime is not commercially available and not FDA-approved for prophylaxis. Use thus requires reconstituting from powder or reliable compounding, coupled with proper storage and care to avoid contaminants.George While 47% of surveyed U.S. cataract surgeons now use or plan to use IC antibiotics, 69% said they would adopt them if the current accessibility and financial barriers didn’t exist.Chang In India, where IC antibiotic benefit has also been reported, their utilization rate was only 38%, with moxifloxacin preferred due to its better availability.51 Despite the better access to fluoroquinolones, the cost-effectiveness of cefuroxime is still significantly better than alternatives;Linertova and while moxifloxacin is cost-effective it is not cost-saving.Leung Although vancomycin is also believed to be an effective agent, it is known to be associated with hemorrhagic occlusive retinal vasculitis (HORV)80,81,82 and never recommended as a prophylactic antibiotic by the World Health Organization.
Similar to their use before, during, or after cataract surgery, topical antibiotics have consistently failed to show benefit with IVT injections. Although counter intuitive, it is possible that topical antibiotics in the context of IVT injections may actually increase the rate of endophthalmitis. A 2018 systematic literature review assessed pooled estimates of incidence of endophthalmitis following anti-VEGF injections and found that the rate with post-injection topical antibiotic use was 0.09% versus 0.03% without.52 This finding is consistent with additional literature indicating topical antibiotic use before and after IVT injections is ineffective at reducing rates of endophthalmitis and may actually be harmful;26,53,54 moreover, topical antibiotics at time of IVT injections has been shown to significantly contribute to antibiotic resistance.54 Similar to cataract surgery, best practices involve povidone-iodine preparation of conjunctiva and eyelid margins prior to injections, with additional care taken to avoid the needle contacting the eyelid.55
Antibiotic resistance is a growing concern with multiple studies analyzing the resistance of various isolates in cases of endophthalmitis. Most cases of antibiotic resistance occur following the use of topical prophylaxis and rarely following IC prophylaxis.10 Over the past several decades, the rates of MRSA and resistant S. pneumonia have risen significantly, with methicillin-resistance seen in 34.6% to 41.6% of S. aureus endophthalmitis infections.Holland In 2015, the Antibiotic Resistance Monitoring in Ocular Microorganisms (ARMOR) Surveillance Study56 published susceptibility findings from isolates of the most common organisms that had been sent from clinical centers throughout the United States. Using minimum inhibitory concentration cutoffs, they found methicillin-resistance was common in Staphylococcus isolates, some having multidrug resistance.56 This was similar to non-ocular trends and overall resistance did not change from the previous five-year study period.56
It has been consistently found that all or nearly all 100% of Gram-positive endophthalmitis isolates are susceptible to vancomycin.6,7,9,56 Vancomycin-resistant organisms, such as VRE, have only been found in extremely rare cases of endophthalmitis.7 In one study on Gram-positive endophthalmitis, of the four cases out of 448 that were resistant to vancomycin, all were susceptible to ciprofloxacin.57 Pseudomonas is the most likely Gram-negative organism to be resistant to both ceftazidime and amikacin, but alternative treatments include imipenem and fluoroquinolones.58
Amikacin provides good coverage for CoNS (95.3%) and Gram-negatives (95.7%) but notably weaker coverage of S. aureus (75%) and Streptococcus (8%).9 Ceftazidime provides similar coverage to amikacin for Gram-negatives, worse coverage for CoNS and better coverage for other Gram-positives.9 Currently, the preferred empirical antimicrobials for suspected infections are IVT vancomycin, ceftazidime and voriconazole.12 Specifically for bacterial coverage, spectrum and sensitivities support the empiric use of vancomycin and ceftazidime.11,12 In these known cases of endophthalmitis, the small risk of HORV is significantly outweighed by the benefit of providing swift effective treatment. Emphasis needs to be made that vancomycin is recommended for the treatment of known or suspected cases of infectious endophthalmitis but not for prophylaxis. Routine infection prophylaxis in ocular surgery with vancomycin – in accordance with judicial antibiotic stewardship - should be abandoned and discouraged given the unfavorable risks of resistance and HORV.
Bacillus endophthalmitis, which is more commonly associated with traumatic injury, is especially virulent and may warrant an adjusted strategy when suspected. A case series of 86 Bacillus isolates were all susceptible to gentamicin, with one resistant to ciprofloxacin and five resistant to vancomycin.59 A similar variation in strategy can be seen in India where Gram-negative endophthalmitis is more common and their susceptibility remains higher for amikacin than for ceftazidime.58
The Endophthalmitis Vitrectomy Study,6 published in 1995, recommends IVT antibiotics to cover both Gram-positives and Gram-negatives in patients who present with vision of hand motion or better, as there was no outcome difference between initial tap-and-inject alone versus immediate vitrectomy. If patients were diabetic or had vision of light perception or worse, vitrectomy was indicated for optimal outcomes.6 For decades, the standard management of infectious endophthalmitis has remained the injection of IVT antibiotics. As we presented above, intravitreal vancomycin and ceftazidime are currently the most common empirical choices. Amikacin, which was used in the original EVS, has been associated with retinal toxicity risk and is no longer routinely used in most places.60 Empirical coverage for fungal sources is uncommon due to the low incidence.61 In cases of known or suspected fungal endophthalmitis, voriconazole or amphotericin B is recommended.18
Yannuzzi et al found that over the past decade, PPV for treatment of post-cataract endophthalmitis decreased in use, from 26% over the prior decade to only 10% of cases.7 Tap-and-inject is widely adopted for primary management in the remaining 90% of cases, with concurrent dexamethasone in 79%.7 This decrease may be due to patients presenting earlier in the disease course with less severe vision deficits, as well as some deviation from previously used strategies. Interestingly, this report shows visual acuity (VA) outcomes were worse during this time period compared to historical literature albeit this was not statistically significantly.7 A recent survey of current strategies in suspected bacterial endophthalmitis found that most respondents adhere to EVS guidelines and generalize them to all types of endophthalmitis.65 PPV was used by 75% of respondents when presenting VA was LP or worse and by 69.8% of respondents when clinical signs worsened after 48 hours of treatment.65 IVT vancomycin and ceftazidime were used by 100% and 96% of the respondents,65 respectively, showing almost uniform implementation of this dual coverage.
A common clinical question facing providers is regarding reinjecting or switching antibiotics after an initial tap-and-inject. Unfortunately, there are only vague recommendations to help govern this decision. It is common for eyes to appear clinically worse after the first day, and thus close monitoring for subsequent improvements by 48 hours is necessary.3 Many patient- and antibiotic-specific factors affect the rate of clearance of medications from the eye, making the timeline complex to decipher. The EVS recommended reinjecting after 36-60 hours if infections were still getting worse, at which point culture and sensitivity information may be available to help tailor treatment. A compendium of published studies showed that at recommended dosage levels, ceftazidime should be reinjected at 48-72 hour intervals and vancomycin at 72 hour intervals.83
There does not exist a clear set of guidelines for treating endophthalmitis following IVT injections, unlike those put forth for post-cataract surgery cases. The similar microbiological profile of IVT-injection endophthalmitis compared to cataract surgery has guided most to employ post-cataract antibiotic guidelines.18 The two common approaches are tap-and-injection of IVT antibiotics alone, or PPV in conjunction with IVT antibiotics. Recently, our retrospective study showed no difference between initial tap-and-inject versus immediate vitrectomy for patients developing endophthalmitis after anti-VEGF injection.19
There are no official guidelines for the treatment of BRI, which has led to a wide variety of approaches. It is important to promptly treat and monitor any bleb-related inflammation in hopes of averting progression to BRE.39 Given the differing microbiological profiles, it would be inappropriate to routinely implement EVS recommendations that were adapted from post-cataract cases.39 Razeghinejad et al assert that using the EVS guidelines may not optimally treat BRE but it could be considered in early-onset cases.38 Among implanted patients who develop endophthalmitis, GDI erosion is common and removal is often needed.36
While EVS provides clear guidelines for the management of post-cataract surgery endophthalmitis, there have been a multitude of updates since its introduction. Surgical techniques have evolved but a lack of prospective randomized controlled trials to provide concrete evidence-based recommendations has limited adoption. Here we discuss novel approaches for the management and treatment of infectious endophthalmitis.
The gold standard for microbial diagnosis remains vitreous sampling and culture. Sampling from the anterior chamber (AC) is sometimes included, but is of limited value because of increased contaminants.60 Vitreous sampling can be done either through small-needle manual aspiration or intraoperatively during PPV. The methods appear to have significantly different yields, favoring PPV.19 Several studies have found culture-positive rates in IVT-related endophthalmitis in the 40% to 60% range.19,23,66 Given the stagnant or possibly worsening yields of microbial culture since the EVS was published, where 69.3% positivity was reported, molecular diagnostics such as polymerase chain reaction (PCR) and deep sequencing of DNA has been explored as a more sensitive testing alternative.67 The improved sensitivity may be of particular utility in cases of chronic endophthalmitis, where slow growth can cause deficient detection.68 Unfortunately, there is also an increased risk of contaminants confounding results, and its usage is limited by access and lack of control standards.3
Despite oral corticosteroid use falling out of favor since the publication of EVS, intravitreal steroids have become common in the acute management of endophthalmitis. The use of steroids remains divisive, as there is very little randomized-controlled trial data, and data that does exist is inconsistent to draw clear conclusions about efficacy or complication rate versus antibiotics alone.62 While they’re used with the intention of controlling collateral damage from the inflammatory response, concern exists over compromised infection control as a possible tradeoff. This fear centers primarily on fungal infection.
Multiple retrospective studies looking at adjuvant IVT steroid efficacy have shown no improvement in visual outcomes or inflammation over non-steroid groups.63,84 Similarly, a recent prospective placebo-controlled superiority trial64 also found that adjuvant IVT dexamethasone did not improve VA outcomes. Multiple confounding factors exist in the available literature including heterogeneous patient populations and whether or not they received initial PPV. At present, evidence does not seem to lend strong support for or against IVT steroids in acute endophthalmitis management. The body of available conflicting data tends to show that IVT steroids with antibiotics are at least no worse than antibiotics alone, and more than likely have a short-term effect. The theoretical risks of using IVT steroids also appear to lack concrete data, making their use a reasonable consideration, especially when the inflammatory reaction to infection is robust.
Limiting vitrectomy to VA of LP or worse at presentation has been the standard protocol since EVS publication; however, debate now exists on whether this is still a valid recommendation. PPV when VA was better than LP has recently shown positive functional and anatomical outcomes.18 The original EVS utilized a three-port 20-gauge PPV technique16 that has been used as a basis of comparison for newer widely adopted methods. Additionally, when performed, the EVS did not recommend complete vitrectomy for fear of complications such as retinal tears.16 Current approaches with earlier complete vitrectomy,69 including removal of posterior cortical vitreous and vitreous membranes, likely leave less bacterial and toxin loads. The presence of newer microsurgical techniques that are safer and provide better long-term outcomes69 could be grounds for updating the guidelines to broaden the use of PPV acutely. Other advances in vitreoretinal surgery include improved illumination, wide-angle viewing, corneal limbal vitrectomy techniques,70 and intraocular endoscopes. The development of new microsurgical instruments has also led to using a valved small-gauge vitrectomy cannula for vitreous tap-and-inject, improving on patient comfort in an office setting.78,79
Several studies have looked at the visual outcomes, microbiology yields, and safety of small-gauge vitrectomy (i.e. 23-, 25-, and 27-gauge) compared to standard 20-gauge. Small-gauge instrumentation allows for less surgical trauma, self-sealing sclerotomy wounds that can be left open, shorter surgical times, decreased corneal astigmatism, and improved patient comfort.71,72,Spirn Self-sealing sclerotomy complications include wound leakage and hypotony, which itself is a risk factor for endophthalmitis.Spirn,Takashina Despite its current use in practice, the data comparing small-gauge to 20-gauge PPV in endophthalmitis management is still limited. We previously looked at retrospective data on 61 cases of infectious endophthalmitis, 27 treated with 20-gauge and 34 treated with 23- or 25-gauge. There was no significant difference seen between groups in the achievement of final VA of 20/40 or better, but significantly more 20-gauge biopsies had culture-positivity.71 Based on this observation, we also investigated the hypothesis that a smaller lumen and increased cutting speed contributed to lower culture-positive rates. This was done by using in vitro cultures with normalized counts of S. epidermidis for aspiration with the different instrument gauge sizes and cutting speeds. Ultimately no significant difference was found, leaving a need for further investigation into the lower culture-positivity with small-gauge systems. Nevertheless, we concluded that small-gauge vitrectomy is a viable option for treatment of infectious endophthalmitis, with comparable visual outcomes to 20-gauge.71 Another study looking at 23-gauge transconjunctival sutureless vitrectomy for acute postoperative endophthalmitis in Taiwan concluded that it was a safe and effective option with good visual outcomes.73 Reduced wound leakage and hypotony can potentially be achieved by using an oblique incision or multi-stage incision similar to cataract surgery.Lin,Shimozono
The use of even smaller instrumentation has shown positive results in multiple retrospective studies. BRE treated with 27-gauge vitrectomy is viable with good safety.74 The increased morbidity and virulence seen with BRE makes safe and effective treatment options of great value. As mentioned earlier, the benefits of these small-gauge approaches include shorter surgical and recovery times, with less conjunctival damage71,72 that might otherwise complicate future glaucoma surgeries.74 Added benefits of 27-gauge instrumentation may include even better wound integrity and elimination of the two-stage incision adopted for 23- and 25- gauge vitrectomy.72
In an effort to address the lack of management guidelines for IVT injection-related endophthalmitis, we performed a retrospective study to compare the outcomes of immediate tap-and-injection (TAI) to initial PPV with IVT antibiotics.19 There was no significant difference seen in the primary outcome of best-corrected visual acuity (BCVA) at 6 months; however, significantly higher culture positivity was seen in the PPV group (90.9%) compared to the TAI group (48.3%).19 Both of these findings are likely influenced by selection bias that saw more patients with advanced disease and poorer presenting VA treated with PPV. Although differences between groups at presentation did not reach statistical significance, a preference towards PPV in advanced disease likely increased culture-positivity while limiting 6-month BCVA in the PPV group.
Inflammation of the anterior segment and media opacity leading to poor visualization can be a barrier to an early, safe and effective PPV. A novel approach we advocate uses a combined five-port limbal and pars plana vitrectomy to remove AC debris for improved visualization prior to vitreous removal.70 After washout of the debris in the AC, the infusion line could be placed appropriately via a pars plana cannula. In our recent case report, we utilized 25-gauge instrumentation that was seen to be safe and free of complications.70
Endoscopic visualization of the interior of the eye has shown benefit with intraocular procedures for many years. Ever since 1990, 20-gauge systems have enabled improved visibility for surgeons in cases of anterior vitreoretinal pathology.75 Surgical treatment of severe endophthalmitis using 20-gauge endoscopy-assisted vitrectomy with the Endo Optiks E4 system was shown to be a useful technique that helped with control of infection and recovery of VA in a small group of patients.76 The more recent miniaturization of instruments has bolstered interest in 23-gauge endoscopes during vitrectomy when visibility is poor secondary to anterior segment opacity. Although they sacrifice some image resolution for their smaller size, they are still adequate for use in endophthalmitis patients.75
Infectious endophthalmitis is a rare but devastating complication of ocular surgical procedures with a decreasing incidence over the past several decades. Major strides have been made with the adoption of povidone-iodine into standard practice during surgery and intravitreal injections. Nevertheless, there are still significant knowledge gaps in our understanding of optimal management strategies. Microorganism identification through culture-positivity has stagnated at approximately 40% to 60% since the 1980s. As modern and novel approaches to vitrectomy continue to evolve, with improved visualization of the posterior segment, decreased surgical times, and less surgical trauma, there is hope for gains to the surgeon’s armamentarium.