Ocular Infections

Visit the link below to download the full version of this book:

https://medidownload.com/product/ocular-infections/

Click Download Now

Preface

Infections of the eye continue to cause serious ocular morbidity and loss of
vision. It is, therefore, highly desirable to maintain a high level of awareness
on new developments in the diagnosis and management of infectious diseases
of the eye. Both classic infections and emerging infections pose serious
threats to man. Many kinds of ocular infections may stay mysterious and
sometimes are hard to diagnose. As medical scientists, we have the unique
role to dispel the mystery of infections of the eye and replace speculations
with certainty and fiction with fact.
This book is an attempt to provide an update on the topic of infections of
the eye and adnexae and provide simplified information that would keep the
ophthalmologist and the eye care practitioner abreast of the recent advances
in antibiotics and in the management of infectious diseases of the eye and
adnexae. Certain infections have been eradicated by mass vaccinations, and
others have been controlled by public health measures, but the resilience of
viruses and the tenacity of bacteria have led to the changes in the pattern of
ocular infections and led to the evolution of old diseases and the emergence
of newly discovered disorders.
In the past two decades, there have been a number of infectious diseases,
and new pathogens have been discovered. Barry and Marshall received the
Nobel Prize for medicine in 2005 for their discovery of Helicobacter pylori
as a cause of peptic ulcer. Bartonella henselae has been identified as the
cause of Bacillary angiomatosis (cat scratch disease) or Bartonellosis. Herpes
virus type 8 was discovered as a cause of Kaposi sarcoma in patients with
AIDS. Human immunodeficiency virus (HIV) 1 and 2 are now well known to
cause AIDS. T-cell lymphoma is caused by human T lymphocyte virus type 1,
and Lyme disease is, caused by Borrelia burgdorferi. In addition, Whipple
disease became known to be caused by Trophyrema whippelii. Similarly,
severe acute respiratory syndrome (SARS) and Avian influenzae are diseases
caused by Corona virus and H5 N1 virus, respectively and Ebola virus is
emerging as a cause of Ebola fever. Infections can cause diseases by invasion
of tissues and destruction of the architecture of visually important structures
in the eye. Infections may lead to immune dysregulation and trigger immune-
mediated inflammation of the eye.
Strifes and natural disasters come and go, but infections are going to be
with us forever. Certain infections may cause blindness, and therefore, oph-
thalmologists should remain updated on these disorders.

v
vi Preface

This book is divided into 16 chapters, making it an easy reference to find

a rapid source for the management of infections of the eye. The book is illus-
trated by figures showing different ocular pathologies and is loaded with
tables summarizing the findings.
The book is clinically oriented with a focus on prospects of infections of
the eye that are important to the clinicians, including the essentials for diag-
nosis, laboratory workup, and management. The treatment is a simple
evidenced-based approach with detailed information on antibiotics use and
dosages. The chapter on antimicrobial agents in ophthalmology summarizes
the antibiotics that are available commercially for infections of the eye, as
well as the compounded topical medications that are not available in the mar-
ket. Antimicrobial agents are commonly used in ocular infections and are
among the most commonly prescribed drugs in community-based physician
offices and hospitals. The use of these antimicrobial agents are outlined in all
the chapters of this book.
Despite our enormous effort and progress in the quest to conquer infec-
tions, there is significant threat from resistant microorganisms.
The information provided in this book is an update on ocular infections,
placing in the hands of ophthalmologists and health practitioners a shortcut
and an easy access to the management of ocular infectious diseases.

Khalid F. Tabbara, MD, ABO, FRCOphth

Riyadh, Saudi Arabia
Baltimore, MD, USA
Ahmed M. Abu El-Asrar, MD, PhD
Riyadh, Saudi Arabia
Moncef Khairallah, MD
Monastir, Tunisia
Contents

1 Molecular Diagnosis of Ocular Infections . . . . . . . . . . . . . . . . . 1

Jolanda D.F. de Groot-Mijnes
2 Antimicrobial Agents in Ophthalmology. . . . . . . . . . . . . . . . . . 19
Khalid F. Tabbara
3 Infections of the Orbit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Moncef Khairallah and Sonia Attia
4 Infections of the Lacrimal System . . . . . . . . . . . . . . . . . . . . . . . 45
Khalid F. Tabbara
5 Infections of the Eyelids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Moncef Khairallah and Rim Kahloun
6 Infectious Conjunctivitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Khalid F. Tabbara
7 Infectious Keratitis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Khalid F. Tabbara and Charbel T. Bou Chacra
8 Viral Anterior Uveitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Samir S. Shoughy and Khalid F. Tabbara
9 Infections of the Posterior Segment: Ocular Tuberculosis. . . . 103
Ahmed M. Abu El-Asrar, Marwan Abouammoh,
and Hani S. Al-Mezaine
10 Infections of the Posterior Segment: Ocular Syphilis. . . . . . . . 119
Sonia Zaouali, Rim Kahloun, and Moncef Khairallah
11 Infections of the Posterior Segment: Other
Bacterial Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Bechir Jelliti, Imen Khairallah-Ksiaa, and Riadh Messaoud
12 Infections of the Posterior Segment:
Parasitic and Fungal Infections . . . . . . . . . . . . . . . . . . . . . . . . . 135
Moncef Khairallah and Rim Kahloun
13 Infections of the Posterior Segment: Acute Retinal Necrosis . . . 155
Hani S. Al-Mezaine, Marwan Abouammoh,
and Ahmed M. Abu El-Asrar

vii
viii Contents

14 Infections of the Posterior Segment:

Cytomegalovirus Retinitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Marwan A. Abouammoh, Hani S. Al-Mezaine,
and Ahmed M. Abu El-Asrar
15 Emergent Ocular Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Moncef Khairallah, Salim Ben Yahia, and Sana Khochtali
16 Endophthalmitis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Hani S. Al-Mezaine and Ahmed M. Abu El-Asrar
 Molecular Diagnosis of Ocular
Infections 1
Jolanda D.F. de Groot-Mijnes

1.1 Introduction diagnosis of ocular infections, investigation of

ocular fluids and biopsies is recommended.
The rapid identification of ocular infections can Molecular diagnostics, of which the poly-
be crucial since treatment regimens and visual merase chain reaction (PCR) is the most applied
prognosis may be entirely different from nonin- procedure, has taken a prominent position within
fectious disorders. The fast identification of spe- the laboratory diagnostic repertoire, also within
cific infectious agents is particularly important in the field of ophthalmology. Molecular assays
immunocompromised patients [1, 2]. In the clini- can be highly sensitive and specific, and theoreti-
cal practice, the presumed diagnosis of an ocular cally, molecular diagnosis can be performed on
infection is based on specific clinical features; all types of ocular materials, provided the pro-
however, laboratory data may be helpful to con- cedure for a particular pathogen is available. For
firm a suspected diagnosis, since similar clinical many viral infections, PCR analysis has replaced
features might be caused by different infectious the less sensitive culture technique. However, in
agents. Laboratory tests based on the analysis of case of bacterial and fungal infections culture
peripheral blood alone are often of limited value, is still preferred over molecular methods, for
since these may not be informative on ocular pro- instance, in orbital cellulitis, mycotic keratitis,
cesses and positive results may be coincidental and endophthalmitis. Viral conjunctivitis is often
[3–5]. Negative peripheral blood results may ren- a self-limiting disease, questioning the neces-
der a specific diagnosis unlikely but do not rule sity to perform molecular diagnosis even though
out the possibility of infection. For a definitive PCR assay is available for among other adenovi-
rus, herpes viruses, and enteroviruses. However,
in case of bilateral blepharoconjunctivitis in
children, herpes viral PCR analysis of tear fluid
J.D.F. de Groot-Mijnes, PhD
Section Clinical Virology, and aqueous humor may be highly valuable for
Department of Medical Microbiology and a rapid diagnosis [6, 7]. Currently, PCR analy-
Department of Ophthalmology, sis is mostly applied for the diagnosis of ocular
University Medical Center Utrecht, trachoma and infectious uveitis and is taking a
Heidelberglaan 100, 3584 CX, Utrecht,
The Netherlands flight for the diagnosis of (mycotic) keratitis and
e-mail: [email protected] endophthalmitis.

K. Tabbara et al. (eds.), Ocular Infections, Essentials in Ophthalmology, 1

DOI 10.1007/978-3-662-43981-4_1, © Springer-Verlag Berlin Heidelberg 2014
2 J.D.F. de Groot-Mijnes

1.2 Molecular Techniques intraocular infections, including cytomegalovi-

for the Diagnosis of Ocular rus (CMV) retinitis, ocular toxoplasmosis, acute
Infections retinal necrosis (ARN), and herpetic anterior
uveitis [13, 19–22]. Specific PCR assays are also
1.2.1 Polymerase Chain Reaction available for many bacteria, like Bartonella
henselae, Borrelia burgdorferi, Treponema pal-
PCR is a technique whereby with the use of short lidum, Mycobacterium tuberculosis and
complementary DNA fragments called primers Mycobacterium species, and Coxiella burnetii,
and DNA polymerase, a single or few copies of a and have been reported to be applicable for the
part of the target DNA are amplified across sev- diagnosis of uveitis [23–31]. PCR directed to the
eral orders of magnitude, generating millions or 16S rRNA conserved gene sequences of bacteria
more copies of a particular nucleic acid sequence is applied to detect bacteria that cause endo-
[8]. The introduction of the PCR has greatly phthalmitis but may also be useful for the diag-
improved the detection of infectious agents. nosis of uveitis entities [29, 32–35]. The 16S
Results can be obtained much faster than with rRNA gene sequences contain hypervariable
culture. Moreover, particular PCR procedures regions which can provide species-specific
have been designed to be so sensitive as to replace sequences which allow for bacterial identifica-
culture in many cases, most notably for viruses tion. As a result, 16S rRNA gene sequencing has
and fastidious microorganisms [9, 10]. become prevalent in medical microbiology as a
Next to the basic gel-based PCR method, various rapid, accurate alternative to phenotypic meth-
more sensitive and specific techniques are available, ods of bacterial identification [36]. Similarly,
such as nested PCR and real-time PCR. Nested PCR PCR assays have been developed targeting
is a modification of the basic PCR and involves two highly conserved gene sequences in fungi, such
sets of primers, used in two consecutive amplifica- as the 18S and 28S rRNA genes. However, one
tion runs, the second set intended to amplify a target has to be aware of possible contamination with
within the first run product, thereby increasing both nonpathogenic microorganisms, as these may
sensitivity and specificity [11, 12]. accidentally be isolated during sampling and
Real-time PCR is also a modification of the surgical procedures [37].
basic PCR but adds a fluorescent compound to Positive PCR outcomes are directly related to
the reaction allowing for real-time monitoring of the pathogenic load in the ocular sample. It has
the PCR reaction and quantitation of the patho- therefore been suggested that the probability of
genic load of the original sample. In case of detecting viruses by PCR is higher than for bac-
Taqman real-time PCR, the fluorescent com- teria or parasites, because viruses generally pro-
pound constitutes a DNA probe, which in addi- duce more progeny [3, 4, 13]. False-positive
tion to the possibility of quantitation increases results for viruses may occur due to contamina-
the specificity of the reaction. tion of samples, overflow from the peripheral
The introduction of real-time PCR assays in blood into the eye or the (intra)ocular presence of
the clinical microbiology laboratory has led to cells infected with infectious agents unrelated to
significant improvements in the diagnosis of the ocular inflammation [3, 38–40]. Therefore,
infectious disease [13, 14]. Contrary to nested positive PCR findings do not always prove
PCR assays, real-time methods allow rapid DNA causality.
amplification, detection, and quantitation of the False-negative results might occur because of
pathogenic load. Moreover, as real-time PCR a low inherent (intra)ocular pathogenic load or
assays are performed in a closed-tube system, the due to the small volume of ocular fluid available
risk of contamination is reduced [15, 16]. for testing and might also depend on the time
However, real-time PCR assays may be less sen- interval between the onset of infection and sam-
sitive than nested PCR assays [17, 18]. pling. Therefore, negative PCR results do not
PCR, most notably real-time PCR, has proven entirely exclude the presumed diagnosis and
to be valuable for the diagnosis of various other diagnostic tools should be considered.
1 Molecular Diagnosis of Ocular Infections 3

1.2.2 Loop-Mediated Isothermal species and strain identification subsequent to

Amplification (LAMP) PCR analysis. RFLP is a tool by which DNA is
cut by sequence-specific restriction enzymes and
Loop-mediated isothermal amplification or subsequently analyzed on agarose gel. The result-
LAMP is a novel molecular detection tool capa- ing fragments are indicative of a particular patho-
ble of amplifying a few copies of DNA to 109 in gen species or strain. This method is mostly used
less than 1 h at isothermal conditions [41]. The for the identification of particular strains, rather
method uses a set of four primers that produce than for species identification following broad-
multiple stem-loop structures resulting in range PCR assays for bacteria and fungi.
increased amplification of the target DNA as Therefore, RFLP has largely been replaced by
compared to PCR. Moreover, the LAMP primers DNA sequencing in the diagnostic setting. DNA
recognize six distinct regions on the target DNA sequencing involves the actual determination of
adding to the specificity. By introducing a reverse the genetic code of a particular piece of
transcriptase step to make copy DNA from RNA DNA. Sequencing has been largely automated
LAMP can also be used to detect RNA targets and may now allow for high-throughput analysis
[41]. LAMP products can be visualized by aga- of diagnostic PCR products. DNA sequencing
rose gel analysis; however, the high release of following PCR amplification has become the
pyrophosphate results in a turbidity which is vis- standard procedure for pathogen species and
ible by eye. Another improvement is the addition strain identification and antibiotic resistance
of SYBR green, which induces a clearly visible profiling.
color change. The LAMP reaction can also be The DNA microarray technology is a geno-
followed in real time by measuring the turbidity typing tool which allows for the simultaneous
in time or by the addition of fluorescent interca- identification of a wide variety of DNA sequences.
lating dyes, reminiscent of real-time PCR meth- An array consists of a large amount of small spe-
ods, allowing for quantitation of pathogenic load cific DNA probes spotted onto a matrix. PCR
of the initial clinical sample. products are then allowed to hybridize to the
LAMP is performed isothermally which obvi- DNA probes. Due to the introduction of fluores-
ates the need for thermal cyclers, and together cent identification markers coupled to the PCR
with its ruggedness and simplicity, LAMP is products, the arrays are scanned using fluorome-
highly applicable for laboratories with limited try and subsequent data analysis is automated.
equipment or at the point of care by clinicians. Like DNA sequencing DNA arrays can be used
LAMP assays have been developed for the for species identification and antibiotic resistance
molecular diagnosis of several ocular infections, profiling [48].
such as uveitis associated with Mycobacterium
tuberculosis, herpes simplex virus, Chikungunya
virus and West Nile virus, and Acanthamoeba 1.3 Molecular Diagnosis
keratitis [42–47]. However, so far the sensitivity of Trachoma
of the LAMP assays does not exceed that of real-
time PCR [42–46]. Trachoma is the leading cause of blindness
worldwide and is caused by the bacterium
Chlamydia trachomatis, which is also associated
1.2.3 Restriction Fragment Length with venereal disease [49]. Different stains, the
Polymorphism, DNA so-called serovars, have been identified of which
Sequencing, and DNA only A through C cause trachoma, D through K
Microarray cause ocular-genital infections, and the
L-serovars cause lymphogranuloma venereum
Restriction fragment length polymorphism (LGV) [50]. Diagnosis of trachoma by clinical
(RFLP), DNA sequencing, and DNA microarray appearance is poor and molecular testing is com-
are molecular assays which are used for pathogen monly used. Several PCR assays have been
4 J.D.F. de Groot-Mijnes

developed for the detection of Chlamydia tra- caution should be taken as fluorescein and anti-
chomatis, most of them were developed for the biotics, chemical substances frequently used for
detection of genital infections. Only two com- the diagnosis and treatment of keratitis, may
mercial assays have been used to diagnose tra- inhibit the PCR reaction, putatively yielding
choma, the COBAS® Amplicor CT/NG Test by false-negative PCR outcomes [59].
Roche Diagnostics Solutions (Basel,
Switzerland) and the LCx™ by Abbott (Chicago,
IL, USA) [51]. More recently, novel molecular 1.4.1 Viral Keratitis
diagnostic tools have been developed based on
the detection of rRNA targets. The advantage of The most frequent cause of viral keratitis is
rRNA amplification over PCR is that these tests HSV. In addition to HSV, VZV, adenoviruses and
are equally specific, but more sensitive due to the enteroviruses can cause corneal infections [60,
high copy number of the rRNAs [52–54]. The 61]. Clinical manifestations of herpes keratitis
new rRNA tests may prove valuable not only in include epithelial dendritic and stromal keratitis,
detecting low-level infections but may also con- the latter of which is more prevalent in children
tribute to the surveillance of ocular Chlamydia, [7, 62]. Dendritic keratitis is often diagnosed on
particularly in areas mass-treated with antibiot- clinical presentation. Stromal keratitis on the
ics where low-level reinfections are emerging other hand is clinically more difficult to diag-
[55, 56, 52]. nose. The molecular diagnosis of keratitis
involves PCR analysis of corneal swabs or scrap-
ings [7, 63–67]. Due to the location and the
1.4 Molecular Diagnosis nature of the infection, the sensitivity of PCR
of Microbial Keratitis assay appears to be higher for dendritic keratitis
than for stromal keratitis [68]. In case of stromal
Microbial keratitis is an infection and inflam- keratitis or keratouveitis, aqueous humor analysis
mation of the cornea, and due to its progres- may add to the diagnosis [69, 68].
sive and potentially devastating nature, it is A good alternative to corneal scraping and
considered an ophthalmological emergency. swabs is the collection of tear fluid as this is less
Hence, rapid diagnosis of the causative agent is damaging to the already fragile cornea [70–76,
of the utmost importance. In a recent literature 68, 60]. However, the sensitivity of the PCR
review by Karsten et al., 232 different species assay on tear fluid may be lower than on corneal
were found to be involved in microbial keratitis scrapings or swabs; Kakimaru-Hasegawa et al.
of which the fungi represent the largest group demonstrated by real-time PCR lower copy num-
[57]. The most common causes include herpes bers in tear samples after comparing simultane-
simplex virus, varicella zoster virus, the bacte- ously collected tear fluid and corneal scrapings
ria Staphylococcus aureus and Pseudomonas from patients with HSV keratitis [68].
aeruginosa, Acanthamoeba keratitis, and sev- In addition to the diagnosis of keratitis, quanti-
eral fungi among which Fusarium, Aspergillus tative PCR analysis of ocular samples can also be
species, and Candida species [5]. The diagnosis used to monitor reaction to treatment and the
is generally made based on clinical presenta- occurrence of resistance to antivirals [63].
tion and noninvasive imaging methods, comple- Acyclovir (ACV) resistance of HSV has been
mented by microbial investigation, such as in described in keratitis, but so far resistance is
vitro microscopical examination and culture of mostly determined phenotypically, that is, by virus
corneal scrapings and biopsies. In recent years, culture in the presence of varying concentrations
PCR analysis of corneal material has become of ACV [64, 65]. One study described the geno-
increasingly more popular as it provides a rapid typic identification of ACV resistance in HSV iso-
and sensitive tool for the identification of ker- lates from patients with keratitis [77]. Mutations in
atitis-causing infectious agents [58]. However, the HSV genome associated with ACV resistance
1 Molecular Diagnosis of Ocular Infections 5

have been identified and molecular tests for HSV (exogenous endophthalmitis) or from an infec-
genotyping are available [78–81]. Genotypic resis- tion elsewhere in the body through septicemia
tance analysis is faster and more sensitive than (endogenous endophthalmitis) [91]. Infectious
phenotypic resistance analysis as it obviates the endophthalmitis can be caused by bacteria, fungi,
need of cultured virus and may therefore prove parasites, and rarely viruses. Bacterial endo-
valuable for analysis of the small volume ocular phthalmitis predominates, with Staphylococci
samples in the near future. and Streptococci as most common causes, fol-
lowed by fungal endophthalmitis caused by
among others Candida parapsilosis, Aspergillus
1.4.2 Nonviral Keratitis ssp, and Fusarium ssp. [91]. A rapid diagnosis is
often of utmost importance. For the diagnosis of
Although in most institutes the diagnosis of non- endophthalmitis, investigation of vitreous by
viral keratitis is still made by more conventional direct Gram-staining and/or culture is currently
methods, such as clinical presentation and direct mostly applied. More recently, molecular analy-
staining or culture of ocular material, the molecu- sis of ocular fluids is being explored. Large stud-
lar diagnosis of bacterial, fungal, and amoebic ies comparing PCR to classical diagnostic tools
keratitis is gaining interest fast. Molecular analy- are not available yet, but molecular analysis
sis of corneal scrapings and biopsies is fast and clearly has its advantages. PCR can be more sen-
sensitive and allows for rapid identification of the sitive than Gram-staining and culture particular
infectious species [5, 82–84]. in very early stages of the disease, when the caus-
Molecular assays currently described for ative agent concerns a fastidious pathogen or
infectious nonviral keratitis predominantly when the patient has received antimicrobial treat-
include broad-range bacterial and fungal PCR ment [92–94]. Also PCR results are acquired
assays targeting the 16S rRNA gene and 18S and faster than culture outcomes. Moreover, with the
28S rRNA genes, respectively [85–88, 37]. emergence of antimicrobial resistance, also intra-
Microbial species identification can subsequently ocularly, new PCR assays are being developed
be done by sequence analysis of the PCR product which not only identify specific microbes but
or by DNA hybridization procedures [84]. For also determine resistance profiles [95–97].
the identification of Acanthamoeba specific PCR Molecular assays currently described for
assays are available [47, 89, 90]. infectious endophthalmitis mostly include the
Due to the high sensitivity and the broad- broad-range bacterial and fungal PCR assays [94,
spectrum nature of the molecular tests for bacteria 98–102]. In addition, Gram-discriminating and
and fungi, one should be cautious with regard to microorganism-specific molecular assay are
false-positive results caused by commensal non- available [103, 104]. Novel explorations include
pathogenic species [37]. As there appears to be a the use of multiplex PCR techniques and DNA
high concordance between PCR and conventional microarrays, which may allow for simultaneous
diagnostic tools, it has been suggested to reserve detection and identification of a large number of
PCR analysis for those cases where conventional microorganism species and may also include
diagnostics yielded negative results or where a resistance profiling [101, 48].
rapid definitive diagnosis is crucial [88, 5].

1.6 Molecular Diagnosis

1.5 Molecular Diagnosis of Infectious Uveitis
of Endophthalmitis
As the inner eye is a secluded compartment sepa-
Endophthalmitis is an uncommon but sight- rated from the periphery by the blood-retina bar-
threatening intraocular inflammation and can rier, analysis of peripheral blood is often not
originate from a trauma or surgical procedure informative for the diagnosis of infectious uveitis.
6 J.D.F. de Groot-Mijnes

For a definitive diagnosis, the analysis of intra- 1.6.1 Diagnosis of Viral Uveitis
ocular fluid is imperative. To obtain intraocular
fluid for diagnostic purposes, a vitreous or aque- The most common viral ocular infections are
ous tap can be performed [3, 105, 106]. A diag- caused by herpes simplex virus (HSV), varicella
nostic vitrectomy is more aggressive than an zoster virus (VZV), and cytomegalovirus
aqueous tap but yields a larger amount of ocular (CMV). However, over the past decade, new
fluid (0.5–0.7 mL). Possible complications of vit- viruses have been identified as important etiolo-
rectomy are endophthalmitis and retinal detach- gies involved in uveitis, such as rubella virus,
ment; however, their incidence is low [3, 107]. West Nile virus, dengue virus, and Chikungunya
An aqueous tap can be performed in the outpa- virus [120, 121, 117].
tient setting, providing approximately 0.1–
0.2 mL aqueous [108]. This procedure has been 1.6.1.1 Herpetic Uveitis
shown to be safe in the hands of an experienced Herpetic uveitis is an ocular inflammation sec-
ophthalmologist [105, 106]. Various infrequent ondary to viral infection caused by HSV-1 and
complications may occur, such as hyphema, HSV-2, VZV, or CMV. Intraocular herpetic infec-
occurring mostly in patients with a high intraocu- tions may present as anterior (kerato)-uveitis or
lar pressure (IOP) at time of paracentesis [106]. as characteristic types of posterior uveitis, such
To date, no systematic studies have been done to as acute retinal necrosis (ARN), progressive
determine whether vitreous or aqueous is supe- outer retinal necrosis (PORN), and CMV retini-
rior in ocular fluid analyses nor have investigated tis. More recently, atypical and non-ARN types
whether the choice of aqueous or vitreous aspi- of posterior ocular infections with HSV and VZV
rate should be dependent on the location of were reported, as well as hypertensive anterior
inflammation within the eye. However, it has uveitis and Fuchs heterochromic uveitis caused
been reported that aqueous humor analysis pro- by CMV.[122–128, 112] CMV retinitis and
vides a safe and useful first laboratory diagnostic PORN occur predominantly in immunosup-
tool prior to the more perilous vitrectomy and pressed patients, whereas the other entities are
that subsequent vitreous analysis when the aque- prevalent mostly in patients with a competent
ous humor was negative contributes only margin- immune system. Peripheral blood analyses for
ally to the diagnosis of infectious uveitis antibodies against HSV, VZV, and CMV are not
[108–110]. useful, because the majority of adults are sero-
Polymerase chain reaction (PCR) analysis of positive (up to 90, 100 and 90 % worldwide,
intraocular fluids has become an important tool respectively) even without a clear clinical history
for the diagnosis of infectious uveitis and may of disease [129–133]. To establish an intraocular
yield results close to 100 % for acute retinal herpesviral infection without obvious extraocular
necrosis and CMV retinitis [21, 111, 112]. signs analysis of aqueous humor or vitreous fluid
However, in other infectious uveitis entities, PCR is required.
yields may be lower. In these cases, analysis of PCR on ocular fluids has been widely used to
intraocular antibody production by Goldmann- diagnose herpetic posterior and panuveitis [1–4,
Witmer coefficient (GWC) or Antibody Index 13, 129, 108, 19, 134–137, 21, 111, 112].
(AI) calculation may contribute to the diagnosis Particularly in ARN and PORN, PCR analysis
of viral and parasitic uveitis entities [13, 113– has a high success rate. Due to the progressive
119]. The GWC or AI is calculated by determining nature of the infection and the high production of
the ratio of pathogen-specific antibody and total viral progeny, PCR may yield sensitivities up to
antibody, usually immunoglobulin G, in serum 100 % [4, 111, 21, 112]. Other HSV- and VZV-
and ocular fluid thus allowing for the differentia- induced retinitis entities also benefit from molec-
tion between leakage of peripheral antibodies ular diagnosis albeit with a lower yield [122, 128,
into the intraocular fluid and actual local anti- 212]. Possibly, because viral replication occurs at
body production. a lower level or because the viral load has
1 Molecular Diagnosis of Ocular Infections 7

decreased due to antiviral treatment or due to In addition to merely diagnosing a viral infec-
sampling at a later time point in disease [4, 13, tion, PCR analysis of sequential ocular samples
119, 118]. Atypical presentations of posterior may be very useful to monitor the efficacy of
uveitis form a diagnostic challenge, and a delay antiviral therapy particularly when treating dev-
in treatment can be harmful to vision. Here, the astating entities such as ARN, PORN, and CMV
diagnosis may also benefit from quick laboratory retinitis [141, 148–150]. Moreover, PCR analysis
testing of aqueous or vitreous specimens. and subsequent sequence or DNA microarray
The diagnosis of CMV retinitis is usually analysis may also be applied to detect antiviral
based on the typical ophthalmoscopic picture. resistance [151].
However, in an immunosuppressed individual,
the introduction of HAART might influence the 1.6.1.2 Fuchs Heterochromic Uveitis
clinical presentation, rendering the clinical diag- Syndrome and Rubella Virus-
nosis more difficult [138–140]. Also for the diag- Associated Uveitis
nosis of CMV retinitis, PCR analysis of Fuchs heterochromic uveitis syndrome (FHUS)
intraocular fluid may be useful and is highly sen- is a chronic low-grade anterior chamber inflam-
sitive particularly in patients with AIDS [141, mation characterized by typical clinical signs
108, 134–136, 1, 21, 112, 2]. such as fine keratic precipitates, diffuse iris atro-
PCR analysis is also applied for the diagnosis phy and/or heterochromia, the development of
of herpetic anterior uveitis. Sensitivities range cataract, and the absence of posterior synechiae
between 20 and 80 % and are higher when the prior to surgery. In Europe, FHUS was reported
analysis includes multiple herpes viruses, such as to be highly associated with rubella virus as in
combined testing for HSV and VZV [136, 112, almost 100 % of patients intraocular antibody
127, 142–144]. CMV anterior uveitis is a newly production against rubella virus was demon-
recognized entity and includes a range of oph- strated [117, 152–154, 147]. Moreover, Birnbaum
thalmological manifestations overlapping with et al. found that the incidence of FHUS was
HSV, VZV, and rubella virus-induced anterior reduced in patients vaccinated against rubella
uveitis [145, 123, 125, 146]. Chee et al. con- virus, strongly suggesting a causal relationship
firmed CMV anterior uveitis by PCR analysis of between FHUS and rubella virus [155].
aqueous humor; however, other groups also diag- Additional studies showed that rubella virus-
nosed CMV anterior uveitis by GWC analysis associated uveitis represents a range of clinical
alone, suggesting that the sensitivity of PCR is manifestations including true FHUS but also
less than 100 % [124, 126, 147]. FHUS-like entities [156, 146].
In immunocompetent patients, herpes viral Due to the high incidence of natural infection
nucleic acid can be readily detected in the early during the pre-vaccination era and recent vacci-
stages of the disease, whereas at later stages PCR nation programs, the seroprevalence for rubella
assays tend to become negative [4, 13, 119, 118]. virus antibodies is very high (94–96 %) [157].
The most likely explanation is that later in infec- Therefore, serology is not informative for the
tion, the immune system, whether or not sup- diagnosis of rubella virus-associated uveitis and
ported by adequate antiviral therapy, has cleared intraocular fluid analysis is essential. Several
the virus or has reduced the viral load below the reports showed that intraocular antibody produc-
detection level. This may explain the high sensi- tion against Rubella virus is positive in 93–100 %
tivity of PCR for the diagnosis of ARN, where of rubella virus-associated uveitis cases, whereas
paracentesis is usually performed very early in PCR remains negative in the majority of cases
the disease [4]. In immunosuppressed individu- (80–90 %) [117, 152–154]. This may be explained
als, viral replication is hardly or not at all limited by a persistent low-grade infection resulting in a
in the absence of antiviral therapy explaining the low viral load in the aqueous humor [117].
high sensitivity of PCR analysis in CMV retinitis However, FHUS representing a chronic autoim-
in AIDS patients [134, 2]. mune reaction triggered by the virus also remains
8 J.D.F. de Groot-Mijnes

a possibility [158]. In short, contrary to herpesvi- For the diagnosis of OT, detection of anti-T.
ral uveitis, PCR analysis is not recommended as gondii IgG antibodies in peripheral blood is not
a first diagnostic laboratory tool for rubella virus- informative in areas with a medium to high sero-
associated uveitis, but may be considered when prevalence. Detection of serum IgM may be use-
intraocular antibody production is detected ful, if OT accompanies a recently acquired
against rubella virus. Toxoplasma infection [174, 175]. To confirm the
diagnosis of toxoplasmosis, intraocular fluid
1.6.1.3 Other Viruses analysis can be performed to detect T. gondii
Other viruses less frequently associated with DNA by PCR and/or to establish intraocular anti-
uveitis can also be identified by PCR analysis body production by Goldmann-Witmer coeffi-
such as Chikungunya virus, dengue virus, HIV, cient [175]. Immunoblotting has also been
human herpes virus 6, and West Nile virus, described for the detection of serum and intraoc-
although information about the sensitivity of the ular antibody; however, this method is elaborate
PCR assays is unavailable [120, 39, 159–162, and quantitation of specific bands is more com-
121]. Epstein-Barr virus can also been identified plicated [22, 176]. Several studies on PCR analy-
by PCR in ocular fluids of uveitis patients; how- sis of Toxoplasma in aqueous humor reported
ever, its role in uveitis remains controversial [39, positive results ranging from 13 to 36 % [177, 13,
163, 112, 164]. A pathognomonic ocular mani- 178, 22]. Analysis of intraocular antibody pro-
festation has not been identified yet. Moreover, duction reportedly yielded positive results up to
intraocular fluids of patients with laboratory- 93 % and, therefore, appears to play a more deci-
proven Toxoplasma chorioretinitis and VZV- sive role in the diagnosis of intraocular
induced acute retinal necrosis and of patients Toxoplasma infection [13, 178, 108, 22, 2]. In
without intraocular inflammation have been primary OT, both PCR and GWC analysis con-
found PCR-positive for EBV. Therefore, PCR tribute equally to the diagnosis of ocular disease
results should be interpreted with caution and [178, 174, 108, 22]. In immunocompromised
preferably performed in combination with other patients, both assays appear to be valuable; how-
laboratory tests, such as determination of intra- ever, PCR was reported to perform best in atypi-
ocular antibody production [39]. cal toxoplasmic chorioretinitis in these patients
[178, 2, 174].
Various studies suggest that for the diagnosis
1.6.2 Diagnosis of Parasitic Uveitis of OT, the application of both diagnostic assays is
indicated irrespective of the patient’s immune
1.6.2.1 Ocular Toxoplasmosis status [177, 13, 176, 2].
Ocular toxoplasmosis (OT), caused by the
parasite Toxoplasma gondii, is the most com-
mon identifiable cause of posterior uveitis in 1.6.3 Diagnosis of Bacterial Uveitis
many parts of the world and can be acquired
either by congenital or postnatal route of infec- In uveitis, the bacterial load in ocular fluids is
tion [165]. Classically, OT presents as a unilat- usually too low to be detected by culture.
eral focal retinochoroidal lesion [166–169]. In Moreover, the involved bacteria are frequently
immunocompromised patients, OT may exhibit fastidious or grow obligatory intracellular. PCR
a variety of clinical manifestations, including analysis may overcome these problems. Several
single foci of retinochoroiditis in one or both molecular assays exist for the detection of intra-
eyes, multifocal lesions, or diffuse areas of reti- ocular bacteria. Specific assays are available for
nal necrosis, and occasionally as AU [170–172]. Mycobacterium tuberculosis, Treponema palli-
Toxoplasma infection may also mimic ARN and dum, and Tropheryma whipplei [179–181,
should be considered when diagnostic testing for 31, 29]. PCR assays are also available for Borrelia
HSV, VZV, and CMV is negative [173]. and Bartonella; however, positive results on
1 Molecular Diagnosis of Ocular Infections 9

ocular fluid have sporadically been reported [182, The diagnosis of ocular tuberculosis is defini-
29]. For the detection of bacteria, the panbacte- tive when M. tuberculosis is identified in the eye.
rial 16S rRNA gene-based assay may be used However, this is rarely achieved, because myco-
[29]. Identification of the particular bacterial spe- bacterial culture facilities are not readily avail-
cies requires subsequent sequence analysis of the able and cultures may require several weeks for a
amplification product. The latter method has positive result [191, 196]. A rapid procedure for
proven valuable for screening large cohorts of diagnosing tuberculosis is the examination of
undiagnosed uveitis and endophthalmitis patients acid-fast (Ziehl-Neelsen) stained smears of
but may also be of value for individual ophthal- infected ocular tissue or fluid. But, as the amount
mology patients with a strong suspicion of a bac- of microorganisms found in intraocular fluids is
terial intraocular infection for which a specific usually relatively low, direct microscopy of the
PCR assay is not available [35, 29, 99, 183]. smears may not be sensitive enough [197, 184].
M. tuberculosis-specific molecular tests, mostly
1.6.3.1 Ocular Tuberculosis PCR assays, are available and have been found
Ocular tuberculosis may present in many enti- useful for the diagnosis of intraocular tuberculo-
ties including conjunctivitis, keratitis, scleritis, sis using several intraocular materials, such as
anterior granulomatous inflammation, retinal aqueous humor and vitreous fluid, but also cho-
vasculitis, or chorioretinal lesions similar to ser- rioretinal biopsies and subretinal fluid [179, 23,
piginous-like choroiditis. The ability to mimic 180, 27, 198–200]. Recently, also LAMP assays
other infections is in part determined by the were developed for Mycobacterium tuberculosis
variable host response and to the fact that virtu- [45, 46]. Little information is available on the
ally all parts of the eye may be affected [184– sensitivity of the molecular tests for the various
186, 83]. This large variation in presentations ocular tuberculosis entities and the ocular mate-
makes the diagnosis of intraocular tuberculosis rial analyzed. However, anecdotal data suggest
difficult. Clinical suspicion is an imperative first that these assays are most useful in tuberculosis
step toward the correct diagnosis. Patients sus- endemic areas where intraocular bacterial loads
pected of ocular tuberculosis generally undergo may be higher.
a complete physical examination including a
Mantoux tuberculin skin test (TST) and chest 1.6.3.2 Ocular Syphilis
radiograph. However, the TST test results The spirochete Treponema pallidum is the caus-
should be interpreted with care. Vaccination ative agent of syphilis, a sexually transmitted
with the bacillus Calmette-Guérin (BCG) vac- disease [201]. Uveitis is the most common ocu-
cine poses a potential source of cross-reactions lar feature of syphilis and is often associated
and may yield false-positive results [184, 187]. with neurosyphilis [202]. No pathognomonic
Recently, interferon-gamma release assays features exist for syphilitic uveitis and, hence,
(IGRAs), such as the QuantiFERON-TB Gold the term “great imitator” applies not only to sys-
test and the T. Spot-TB® Elispot assay have been temic syphilis but also to ocular syphilitic dis-
added to the diagnostic repertoire [188, 184, ease. When ocular syphilis is suspected, initially
189, 190]. The antigens used in these assays are standard syphilis screening assays are performed,
not shared by the vaccine strain, thus prevent- such as the Treponema pallidum hemagglutina-
ing false-positive results in BCG-vaccinated tion and particle agglutination assays (TPHA
individuals [188, 184]. A positive TST or IGRA and TPPA, respectively), fluorescent treponemal
only indicates that a person has been exposed to antibody absorption (FTA-ABS) test and the
Mycobacterium tuberculosis, and does not dis- immunoblot on peripheral blood. Enzyme immu-
tinguish between active or latent disease. In fact, noassays are also available and show promising
in active disease or in AIDS patients with a low results as screening assays in all stages of syphi-
CD4 count, both the TST and IGRAs may be lis [203, 204]. However, none of these tests dis-
negative [191–195]. criminate between a previous or active infection.
10 J.D.F. de Groot-Mijnes

The non-treponemal Venereal Disease Research On the other hand, PCR assays have become
Laboratory (VDRL) and the rapid plasma reagin so sensitive that occasionally infectious agents
(RPR) tests are applied to determine the activity are detected that are merely bystanders and do
of disease and can be useful to monitor treatment not contribute to the actual disease.
[201]. Neurosyphilis is confirmed by a positive Other diagnostic tools may complement
VDRL in CSF or by the presence of intrathecal molecular diagnostic tools, not only to confirm
antibody production, using the TPHA or TPPA a diagnosis in case of a negative PCR result but
[205–207]. The diagnosis of ocular syphilis is also to confirm or exclude a diagnosis in case
considered proven when ocular disease is present of a putative false-positive PCR result.
in active syphilis or in combination with proven Antibody detection may provide a valuable
neurosyphilis. However, in case of severely immu- addition to the diagnostic repertoire. In kerati-
nocompromised patients, the possibility of other tis and conjunctivitis, IgA detection in tear
ocular infections should preferable be excluded. fluid can be of help [210, 211]. In infectious
Intraocular fluid analysis is not commonly used uveitis, the detection of intraocular antibody
for the diagnosis of ocular syphilis. PCR assays production can play a major role in the diagno-
are available and have been reported positive for sis of herpesviral, rubella viral, and toxoplas-
both aqueous and vitreous fluids; however, sen- mic infections [4, 152, 13, 119, 118, 175, 22].
sitivity data on the molecular analysis for ocular
syphilis are not available [30, 31, 208, 209].
Compliance with Ethical Requirements

Conclusions
Conflict of Interest The author declares that she has no
Most molecular assays are expensive and
conflict of interest.
require well-equipped laboratories. Moreover,
molecular analysis may not always provide an
Informed Consent No human studies were carried out
answer. However, molecular diagnostic tools by the authors for this article.
clearly have their benefits. Particularly PCR
analysis of ocular fluids can contribute sig- Animal Studies No animal studies were carried out by
nificantly to the diagnosis of an ocular infec- the authors for this article.
tion. However, it also has its weaknesses and
limitations. A positive PCR result may depend
on many factors. Indisputably, it requires the References
presence of the culprit in the collected ocu-
lar sample. Depending on the location and 1. Mitchell SM, Fox JD, Tedder RS, Gazzard BG,
nature of the ocular infection, the pathogenic Lightman S. Vitreous fluid sampling and viral genome
detection for the diagnosis of viral retinitis in patients
load in different ocular samples may dif- with AIDS. J Med Virol. 1994;43(4):336–40.
fer. For instance, in case of herpes keratitis, 2. Westeneng AC, Rothova A, de Boer JH, de Groot-
it was demonstrated that the viral burden is Mijnes JDF. Infectious uveitis in immunocompro-
higher in ocular scrapings than in tear fluid mised patients and the diagnostic value of polymerase
chain reaction and Goldmann-Witmer coefficient in
[68]. For the same reason, Chlamydia tra- aqueous analysis. Am J Ophthalmol. 2007;144(5):
chomatis, an obligate intracellular bacterium, 781–5.
is best detected in conjunctival swabs which 3. Bodaghi B, LeHoang P. Testing ocular fluids in uve-
contain infected cells. Negative PCR results itis. Ophthalmol Clin North Am. 2002;15(3):
271–9.
may also be encountered when a patient has 4. de Boer JH, Verhagen C, Bruinenberg M, Rothova A,
already been treated with antibiotics or anti- de Jong PT, Baarsma GS, et al. Serologic and poly-
virals. Furthermore, in subacute or chronic merase chain reaction analysis of intraocular fluids in
ocular diseases, such as herpetic anterior uve- the diagnosis of infectious uveitis. Am J Ophthalmol.
1996;121(6):650–8.
itis and rubella virus-associated uveitis, the 5. Thomas PA, Kaliamurthy J. Mycotic keratitis: epide-
pathogenic load may have dropped below the miology, diagnosis and management. Clin Microbiol
detection level rendering negative PCR results. Infect. 2013;19(3):210–20.