Inflammatory reactions after intravitreal triamcinolone acetonide: possible mechanisms and therapeutic options

Figure 1. Activated macrophages (arrow) exposed in vitro to Kenalog^® were unable to digest and eliminate triamcinolone acetonide crystals.

The macrophages expanded in size, rupturing and releasing proinflammatory cytokines.

Reprinted with permission from Jeffrey Edelman (Allergan Inc., CA, USA).

Figure 2. Clinical histological evidence of inflammation due to the inability of macrophages to digest triamcinolone acetonide crystals.

Seen in an aqueous humor aspirate of a patient with noninfectious endophthalmitis after intravitreal triamcinolone acetonide, showing neutrophil infiltration and a fibrinous reaction, suggesting presence of phagocytic cells in the vitreous.

Reprinted with permission from Chi Chao Chan, MD (National Eye Institute, MD, USA).

The use of intravitreal triamcinolone acetonide (IVTA) has been commonplace in the treatment of vitreoretinal diseases since its initial human application in 1995 [1]. Unlike infectious endophthalmitis, which is associated with any therapeutic agent delivered by intravitreal injection, noninfectious endophthalmitis and pseudo-endophthalmitis are more commonly associated with IVTA injection than other pharmaceuticals [2]. Occurring in as much as 2.0% of patients, the pathogenesis of noninfectious endophthalmitis after IVTA remains controversial [2,3]. The suggested etiologies include the potential effect of endotoxins [4], preservatives in some TA formulations [5] and the dispersion of triamcinolone acetonide (TA) crystals in the vitreous [6]. Alternative preservative-free TA (PFTA) formulations have been investigated in the hope that they may limit the occurrence of inflammatory conditions that mimic endophthalmitis [4,5,7–10]. Herein, we review the current understanding of the pathogenesis of noninfectious endophthalmitis after IVTA and discuss the propensity for noninfectious endophthalmitis among commercially available TA formulations for intravitreal injection.

Endotoxins in TA formulations have been suggested as a cause of noninfectious endophthalmitis following IVTA injection [4]. Animal models have demonstrated that a single intravitreal injection of 0.01–500 µg of bacterial endotoxin can induce dose-dependant endophthalmitis [11]. Furthermore, endotoxins have also been shown to induce the release of proinflammatory cytokines, chemokines and adhesion molecules from human ocular tissues [12]. A similar, acute anterior segment sterile inflammatory reaction was traced to endotoxin contamination of balanced salt solution formulations, and may represent a clinical analogy to noninfectious endophthalmitis following IVTA [13,101]. However, a recent cluster of sterile endophthalmitis following Kenalog^® IVTA (Bristol Myers Squibb, NJ, USA) was reported and vials from the same lots as those patients developing sterile endophthalmitis were negative for endotoxin [4].

Benzyl alcohol preservative has been implicated in the pathogenesis of noninfectious endophthalmitis after IVTA [5]. Preservatives in intrathecal and epidural formulations have been associated with neurotoxic effects [14]. The use of preservative-free materials in this clinical setting has significantly reduced the occurrence of fatal toxic syndrome in premature infants [15]. Moreover, nonclinical studies have demonstrated that the vehicle or preservatives in the most commonly used IVTA formulation (Kenalog 40 mg/ml) have been shown in animal models to induce an acute inflammatory reaction by exerting a direct toxic effect on retinal cells [16,17]. Benzyl alcohol preservative in commercially available TA has demonstrated toxicity in the rabbit eye at concentrations of 0.073% or more (0.022% corresponds to the dose in 0.1 ml of commercially available TA in the USA) [18]. However, in clinical practice, filtering the benzyl alcohol has not significantly reduced the incidence of noninfectious endophthalmitis, with reported rates of 3.5–7.3% in eyes injected with filtered TA, which reportedly removed 90% of the preservatives [4,8].

Dispersion of IVTA crystals may also contribute to the pathogenesis of noninfectious endophthalmitis. Dispersion of exposed TA crystals within the vitreous provides phagocytic targets that activate resident macrophages [19]. However, as the dispersed crystals are not able to be digested, they accumulate in the phagolysosomes of the macrophage, leading to cell death. In vitro experiments have shown that human macrophages exposed to Kenalog expanded to double or triple their normal volume during phagocytosis of TA crystals (Figure 1). These cells subsequently rupture and release proinflammatory cytokines due to their inability to digest and eliminate the particles from the phagolysosomes following their rapid phagocytosis [Edelman J, Pers. Comm.]. Furthermore, clinical evidence from histology of an aqueous humor aspirate of a patient with noninfectious endophthalmitis after IVTA (Kenalog) showed neutrophil infiltration and a fibrinous reaction, suggesting the presence of phagocytic cells (Figure 2) [Chan CC, Pers. Comm.]. A clinical analogue to the TA crystal-induced inflammation is an intra-articular inflammatory flare after injection of poorly soluble hydrocortisone acetate crystals. Alternatively, an injection of a soluble, crystal-free corticosteroid formulation avoids the acute, sterile inflammatory response in the joint [20].

Currently, there are four available options for clinical use of IVTA: Kenalog, filtered TA, PFTA and TA injectable suspension (TAIS; TRIESENCE^®, Alcon, Inc., TX, USA). The most commonly used IVTA off-label formulation in the past was Kenalog, a sterile suspension of TA crystals in an aqueous medium that contains 0.99% benzyl alcohol as a preservative and bacteriocidal agent [17]. Considering that benzyl alcohol may induce inflammation in the vitreous, filtered TA and PFTA have been investigated [4,8,10]. Filtering of the most commonly used TA formulation can remove up to 99.7% of the benzyl alcohol, and a reduced occurrence of endophthalmitis has been reported with this approach [5]. The limitations of filtered TA include the possible inaccuracy of the final TA dosage following the filtration procedure [6], variability in the increase or decrease of benzyl alcohol in the filtered formulation [7], and possible microbial contamination occurring during the filtering procedure. Furthermore, noninfectious endophthalmitis may still occur in a small percentage of patients despite the administration of benzyl alcohol-filtered TA [4].

Preservative-free triamcinolone acetonide has been commercially formulated by compounding pharmacies [8,9]. Unlike filtered TA, PFTA is typically prepared by direct suspension of micronized TA in a delivery vehicle, such as balanced salt solution or hydroxypropyl methylcellulose [10]. Clinical studies have suggested that PFTA may reduce the risk of noninfectious endophthalmitis [8]. Jonas et al. reported no cases of noninfectious endophthalmitis after 1135 intravitreal injections of PFTA [5]. In other studies with PFTA, noninfectious endophthalmitis occurred at a rate of 1.2% after 577 injections [8] and 2.2% after 502 injections [21].

More recently, another PFTA that is US FDA approved for ophthalmic use has become available (TRIESENCE 40 mg/ml). It carries the advantages of PFTA with regard to avoiding preservatives and also providing a known concentration of triamcinolone. The dispersive effect of this PFTA provides a theoretical advantage in enhancing vitreous and membrane visualization during vitrectomy [9]. A recent study reported noninfectious endophthalmitis at a rate of 5.3% (two out of 38 injections) after intravitreal TRIESENCE injection [22]. Large clinical series using this IVTA preparation have not yet been reported.

Another preservative-free formulation of TA, suspended in a hydrogel vehicle, is Trivaris (Allergan Inc., CA, USA). The TA crystals are suspended in a viscous polymer, which theoretically limits free crystal exposure to macrophages in the vitreous after intravitreal injection, but also reduces the dispersion of the drug throughout the vitreous cavity. The lack of crystal dispersion throughout the vitreous after PFTA hydrogel injection may also reduce the transient clouding of vision experienced by some patients after IVTA injection; however, its viscosity precludes its use as a dispersive agent to better identify the posterior cortical vitreous and membranes during vitreoretinal surgery. Trivaris, prepared in a concentration of 80 mg/ml, is FDA approved but not yet commercially available.

Safety data from two, large, randomized, multicenter clinical trials (the Diabetic Retinopathy Clinical Research [DRCR] Network protocol B and Standard Care Versus Corticosteroid for Retinal Vein Occlusion [SCORE]) revealed no cases of noninfectious endophthalmitis after 2009 intravitreal injections of a PFTA hydrogel (Trivaris) [23]. In vitro exposure of human macrophages to this formulation demonstrated significantly reduced expression of proinflammatory cytokines compared with other preparations. More interestingly, intravitreal injection of this PFTA hydrogel in rats demonstrated suppression of ocular IL-7, an inflammatory cytokine, compared with Kenalog and saline injection [Edelman J, Pers. Comm.].

The various formulations of TA possess significant differences in aggregate size, with Triesence>Kenalog>PFTA [24]. Furthermore, an inverse relationship exists between spectral photometry absorption and aggregate size. In the in vitro setting, this is translated into greater dispersion and less precipitation over time for the larger aggregate sizes. In addition, PFTA demonstrated a larger mean number of aggregates [24]. Ultimately, the impact of particle size, aggregate size and number of aggregates upon predisposition towards noninfectious endophthalmitis remains to be determined.

Enthusiasm for the off-label use of IVTA for vitreoretinal diseases is being tempered by an increased awareness of the potential for ocular complications, including glaucoma and cataract, as well as the occurrence of noninfectious endophthalmitis. The pathogenesis of noninfectious endophthalmitis after IVTA remains controversial, but may involve contributions from multiple mechanisms including endotoxins, preservatives or dispersion of TA crystals. Further investigations and recognition of these contributing factors may influence the development of better IVTA formulations for ophthalmic use.

Expert commentary

Although any eye is at risk of developing noninfectious endophthalmitis after IVTA injection, some eyes pose a greater risk. Eyes that are prone to inflammation to begin with, such as eyes with uveitis or postoperative cystoid macular edema, carry a greater risk of this complication. Newer formulations of PFTA appear to reduce the risk of noninfectious endophthalmitis, and further study of the relevance of triamcinolone particle size in inducing intraocular inflammation will further enhance our understanding of this subject.

Five-year view

Within the next 5 years, more data will be collected with regard to the incidence of noninfectious endopthalmitis after intravitreal administration of different formulations of TA and the factors predisposing to this complication. Furthermore, a better understanding of the pathophysiologic mechanisms leading to the development of inflammation after IVTA will better guide us as to avoiding this untoward event.

Financial & competing interests disclosure

Supported in part by Research to Prevent Blindness, NY, USA. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.