The metabolic end products and microbial virulent factors initiate inflammatory process in the cellular components of the eye leading to hypotony [12, 13]. The retinal tissue damage is irreversible and results in vision loss. Callegan’s experiments further showed the rapid deterioration of retinal responses (electroretinogram, ERG) along with the exponential growth of bacteria in the vitreous humor within 12 h in B. cereus-infected eyes. Pseudomonas is reported to have a longer doubling time of ~3.5 h [14]. Studies using experimentally injected endotoxin of E. coli showed that the vitreous loaded with E. coli endotoxins cleared over 2 weeks [15]. This implies that the inflammatory damage caused by the endotoxins can persist in the eye for prolonged period, even after the microbial kill. Novosad et al. [16] reported elevated levels of cytokine and chemokine along with the involvement of toll-like receptors (TLR2) in infection of B. cereus endophthalmitis; TLR is integral to recognizing the invading organisms causing endophthalmitis. We do not know much about the initiation and extent of alteration of barrier function in terms of transporter susceptibility or paracellular transport of therapeutic agents in the entire cascade of events in endophthalmitis.

Basically there are three models. The antimicrobial agents such as penicillin follow pharmacodynamic model of T > MIC (time > minimum inhibitory concentration), where the concentration-dependent antimicrobial activity occurs over a narrow range of drug concentrations and the extent of antimicrobial activity is a function of the duration of effective exposure. Agents altering protein synthesis such as streptomycin follow C _max/MIC (maximum concentration/minimum inhibitory concentration) model, which shows concentration-dependent bactericidal effect over the range of drug concentration. Agents like vancomycin follow AUC/MIC (area under curve/minimum inhibitory concentration) model.

Apart from their direct action, postantibiotic effect is an additional parameter for an antimicrobial agent in maintaining its activity. Postantibiotic effect is the time period beginning after the organism is exposed to an antimicrobial agent until the survivors begin to multiply to a significant degree. A significant postantibiotic effect is reported with agents inhibiting protein or nucleic acid synthesis such as macrolides, quinolones, tetracyclines, etc. [1]. Postantibiotic effect has been best correlated with the aforesaid PK/PD models of the known classes of antimicrobials [17].

In the Endophthalmitis Vitrectomy Study, a prospective study of post-cataract acute endophthalmitis, 94.2% of isolates were gram-positive pathogens [18]. Most of the antimicrobials approved for human use are for systemic use and are seldom studied for infections in the eye that is well protected by blood-ocular barriers. Systematic studies on various degrees of inflammation affecting the barrier function that in turn alters the pharmacokinetics of antimicrobial agents with differential susceptibility for transporters are not yet explored. The type of antimicrobial agent used and its frequency of dosing would determine the duration or the extent of its exposure leading to effective microbial control. Therefore, applying rational approaches with modified PK/PD assumptions is required to predict the intraocular penetration of antimicrobials in endophthalmitis of bacterial and fungal origins. Considering the complications and restrictions associated with the sampling techniques of ocular fluids, animal studies are often extrapolated to human use. Despite these studies, lack of appropriate data of plasma and corresponding intraocular level (anterior chamber fluid and/or vitreous) for an antimicrobial agent in normal and inflammatory conditions is a hurdle while applying PK/PD assumptions in endophthalmitis (Fig. 23.2).

Factors influencing ocular disposition of antimicrobial agents are conventionally discussed by the route of elimination, ionic nature, solubility coefficient of the drug, status of ocular inflammation, surgical status of the eye, molecular weight, vitreous liquefaction, solution density, and frequency of intravitreal administration [6, 19]. Increasing knowledge of the barriers of the eye has shown new inputs regarding the fate of ocular drug concentration based on their transporter susceptibility [2, 20–22]. Most of the antimicrobials are substrates of anionic, cationic, P-gp (P-glycoprotein), or PEPT (peptide) transporters as per their charge and molecular structure or characteristics [3]. This property determines their effective concentration in the ocular tissues after direct or systemic administrations.

The ocular kinetics of intravitreal gentamicin, eliminated by the anterior pathway, has been explained by Barza et al. in 1983 [23], and recently Nirmal et al. [22] described the elimination of intravitreally injected positively charged compounds based on the functional studies on the position of organic cation transporters (OCT) in rabbits. OCTs are present in ocular tissues in the intake position from the blood and the vitreous; their elimination is achieved through slow anterior pathway. Unlike substrates of P-gp eliminated through retinal pump mechanisms [24], intraocularly injected positively charged compounds follow relatively slow anterior elimination pathway. Therefore, positively charged compounds like ceftazidime, cefepime, and daptomycin (Fig. 23.3) are best suited for intraocular administration as they are unlikely eliminated through the retina, which is very rapid considering the surface area and orientation of transporters. Studies with the help of gamma camera have shown the retinal elimination of radiolabeled ofloxacin (P-gp substrate) via the retina by posterior drug efflux pathway [21]. This study has also shown that ofloxacin’s vitreous elimination could be delayed by blocking the P-gp using verapamil.

In the blood-eye interface, we consider two important barriers involved in the regulation of the exchange of endogenous compounds. The ciliary body and iris contribute toward the blood-aqueous barrier where influx of compounds from the blood into the eye predominates. The blood-retinal barrier regulates the major control over the inward and outward movement of compounds from the retina/vitreous [2]. Blood-retinal barrier is further divided into two types, the inner blood-retinal barrier (iBRB) and the outer blood-retinal barrier (oBRB). The continuous endothelial cell linings of the blood vessels of the neural retina form the iBRB; it rests on the basal lamina that is covered by processes of Muller glial cells and astrocytes [2]. Outer blood-retinal barrier is constituted by retinal pigment epithelium and choroid. These are tightly regulated structures; transport of nutrients plays a physiological role under normal conditions, helping in retinal homeostasis through specific transporter mechanisms (Fig. 23.4).

Blood-aqueous and both retinal barriers of the retina (iBRB and oBRB) express transporters of physiological relevance that are involved in the transport of xenobiotics. The transporters relevant to ocular pharmacokinetics of drugs belong to variety of drug transporters in the ATP-binding cassette (ABC) and solute carrier (SLC) families. Among them P-gp, breast cancer resistance protein (BCRP), multidrug resistance-associated proteins (MRPs), organic anion transporters (OATs), organic anion transporting polypeptides (OATPs), bile acid transporters (ASBT and NTCP), OCTNs and MATE, and peptide transporters (PEPTs) have a potential role in ocular kinetics of drugs injected systemically or intraocularly [2, 20, 25]. Many systematic studies have shown the functional importance of these transporters in the influx and efflux of xenobiotics in blood-ocular barriers. OCT is functionally active in blood-ocular barriers and is involved in the transport of its substrate from the blood to vitreous humor and also in the uptake position in the cornea [26, 27]. P-glycoprotein transporters are involved in the efflux of intravitreally injected substrates, and they also block their systemically administered substrates from reaching adequate concentration inside the eye [28, 29]. Nutrient transporters like peptide transporters (PEPTs) are targeted for the enhanced uptake of peptide-based prodrug across blood-ocular barrier [30].

Presence and function of organic anion transporters in the retinal elimination are known from the studies with intravitreal carbenicillin (OAT substrate) whose vitreous half-life was prolonged when treated with OAT inhibitor probenecid [23]. The presence of equilibrative nucleoside transporter (ENT) has been found in the retina [31] though their role in the ocular pharmacokinetics of drugs is not known. Although the presence of other transporters in the retina has been shown, their functional importance and alteration in various ocular pathological conditions remain unproven. As of now, except positively charged substrates like gentamicin, most of other compounds are reported to follow posterior elimination pathway through the retina. Therefore, the net result of inflammation induced in endophthalmitis on the expression and function of transporters in blood-retinal barrier is a matter of interest to understand the degree of antimicrobial drug influx after systemic administration or its elimination after intravitreal injection.

A meta-analysis on prophylactic use of antibiotics for prevention of endophthalmitis did not show a valid rationale; rather, a strict maintenance of aseptic conditions for ocular surgical procedure remains the gold standard infection prevention [4]. Ocular disposition of antimicrobials after intravitreal injection in normal and inflamed eyes has been extensively investigated in several animal and human studies [19, 32]. But, the rationale for the dose and frequency of repeat injections based on the pharmacodynamic model and toxicity is not available.

There are very few reports that have investigated vitreous concentrations of antimicrobial agents in inflamed and normal eyes following systemic therapy. They do not help us understand the impact of inflammation-induced alteration of blood-ocular barriers. Rajpat et al. [33] have studied intraocular penetration of gatifloxacin after systemic therapy; they showed differential drug penetration into vitreous after oral administration of 400 mg in the patients undergoing vitrectomy in inflamed and non-inflamed circumstances. In the inflamed eye group, vitreous to plasma ratios at 2, 4, and 6 h after oral administration were 0.14, 0.27, and 0.28 in inflamed eyes and were 0.07, 0.21, and 0.26 in control (uninflamed) eyes, respectively. At a dose of 800 mg of gatifloxacin, this study documented an increase of 12.6% of drug concentration in an inflamed eye. Ferencz et al. [34] studied ocular penetration of vancomycin in patients undergoing vitrectomy for endophthalmitis after 1 gm intravenous injection. This study showed that vancomycin levels in vitreous increase with time to 2.04 ± 1.2 μg/ml in 4–5 h after the injection. The vitreous to serum ratio at 4–5 h after the intravenous injection was 0.16 in the infected eyes, not adequate enough for the expected antimicrobial activity [34]. Thus, one could conclude that infection-induced inflammation alters the barrier functions, though it cannot be relied upon as a dependable parameter.

Intravitreal injected antibiotic clearance in inflamed and uninflamed eyes has been studied by Meredith et al. [32] for the amikacin clearance via anterior route and by Ficker et al. [35] for cefazolin clearance via the posterior route. Table 23.1 lists the vitreous pharmacokinetics of intravitreally injected antibiotics in rabbits adopted from Meredith et al. [32], Ficker et al. [35], and Khamdang et al. [36] that has documented the effect of inflammation on the elimination pathways. It is apparent from Table 23.1 that inflammation and aphakia impact the drug elimination from the vitreous cavity. In a rabbit eye, inflammation and aphakia affect differently for the posterior route eliminated drug (e.g., cefazolin) and anterior route eliminated drug (e.g., amikacin, gentamicin). The vitreous concentration of cefazolin is higher and amikacin is lower in inflamed and aphakic eyes compared to normal phakic eyes. Keeping these facts in mind, antibiotic administration schedule must be optimized for maximal effect.