INTRODUCTION

Infectious keratitis occurs worldwide, and despite recent develop ments, it remains a potentially blinding condition⁽¹⁾. Often, a specific diagnosis is not clinically possible. Early diagnosis and immediate appropriate treatment can reduce disease morbidity^(2,3). Laboratory iden tification is important because different etiological agents can manifest themselves with similar clinical profiles, such as Acanthamoeba spp., herpes viruses, bacteria, and fungi^(4-6).

The development of nucleic acid amplification tests, such as the polymerase chain reaction (PCR), has provided more specific and sensitive methods for diagnosis^(7-9). Quantitative real-time PCR (qPCR) has advantages over conventional methods because it is quicker and more sensitive⁽¹⁰⁾. It can also detect genes that are expressed only in the replication state and can relatively quantify the DNA sample^(11,12).

The current study assesses the presence of herpes simplex virus-1 and -2 (HSV-1 and -2) and varicella zoster virus (VZV) by qPCR in corneal scrapings from patients with infectious keratitis. Risk factors are also studied.

METHODS

Patients with clinical diagnoses of infectious corneal ulcers referred for diagnostic corneal scrapings were included in this study (the study group). Risk factors and epidemiological information, as well as the duration of disease and its clinical evolution, were recorded for each patient. All patients signed an informed consent form after approval by the Ethics Committee (1422/06).

All patients underwent a clinical examination that included slit lamp biomicroscopy and photography. Diseased corneas were scraped and specimens were sent for direct examination (Gram and Giemsa staining techniques) and microbiology investigation for bacteria, fungi, and Acanthamoeba as well as qPCR for viruses. At the time of examination, the following information was recorded for each patient: gen der, age, occupation, history of ocular trauma, previous intraocular surgery, presence of local or systemic risk factors (such as contact lens use, diabetes, HIV, advanced age, and corneal exposition), current medications, and symptom duration.

Corneal scrapings were obtained from the edge of the ulcer after application of topical anesthesia and the ocular surface was washed with physiologic saline to avoid false-positive results during qPCR⁽⁹⁾. A portion of the sample was quickly frozen (-80°C) and sent for molecular diagnostic techniques. Cultures were considered positive when there was heavy growth on stria blood or chocolate agar and a positive correlation with the results of smears.

Total DNA extraction was performed using a QIAamp DNA Mini Kit (Qiagen, Valencia, CA), according to the manufacturer's protocol.

Molecular detection was carried out using TaqMan-based techno logy, and reactions for HSV-1 and -2 and VZV (sequences of primers and probes are listed in Table 1) were tested for analytical specificity using positive (known viral DNA from HSV-1 and -2, and VZV) and negative (molecular grade water) controls⁽¹³⁾. The reaction was standardized with primers and probes for HSV-1/2 in a duplex format and a VZV target in a separate reaction. Briefly, 5 mL of DNA was added to a final reaction volume of 25.0 mL (12.5 mL of the 2X TaqMan^TM universal PCR master mix, 0.3 mM of each primer, 0.3 mM of each probe, and 5.7 mL of molecular grade sterile water), and 45 PCR cycles were run: denaturation step (20 sec at 95°C), primer annealing and extension (1 min at 60°C). The beta-globin gene was used as the endogenous control⁽¹⁴⁾ (10.0 mL of 2X TaqMan^TM universal PCR mix, 0.4 mM of each primer, 0.2 mM of the probe, 7.0 mL molecular grade sterile water, and 2.0 mL of DNA sample to a final volume of 20.0 mL) and the same PCR conditions described above were used. All reactions were carried out in an ABI Prism 7500 device (Applied Biosystems, Carlsbad, CA, USA). The clinical samples were then analyzed for the presence of HSV-1 and -2 and VZV DNA. The molecular test was designed to detect only actively replicating virus, through collection of corneal scrapings only from symptomatic patients. The alpha subfamily of the herpes viruses replicates efficiently in skin and mucosa, in which the mucosa sore is a known site of active replication⁽¹⁴⁾. The control group comprised 25 eyes with typical herpes dendritic keratitis that were also analyzed by qPCR.

RESULTS

Sixty-five eyes of 65 patients (38 females; mean age 44.9 years, range 8-93 years) who had been clinically diagnosed with bacterial keratitis (the study group) and 25 eyes of 25 patients presenting with typical herpes dendritic keratitis (the control group) were tested from May 2008 to December 2010 (Table 2).

From the study group, 9 of 65 eyes (13.8%) had negative smears, cultures, and PCR findings. From these nine patients, eight had corneal ulcers <2 mm and one had a 3.5 × 3 mm ulcer. Six patients were contact lens wearers and two reported previous ocular trauma. All eyes with negative cultures were also negative for viruses by qPCR analysis.

The most frequent etiologic agent was coagulase-negative Sta phylococcus spp., in 21 of 56 culture-positive eyes (37.5%). From the culture-positive eyes, gram-positive microorganisms were responsible for the infections in 37 eyes (66.1%). Gram-negative microorganisms were present in 13 eyes (23.2%) and fungi were present in four eyes (7.1%). One eye (1.8%) presented Acanthamoeba growth on cultures. Most of the patients with positive cultures had ulcers >2 mm with variable clinical aspects. Risk factors for keratitis were present in 21 eyes (37.5%) and included trauma, previous intraocular surgery, bullous keratopathy, contact lens wear, and Steven-Johnson syndrome.

In 5 of 56 eyes with positive cultures, more than one etiological agent was detected: two eyes presented two Gram-positive organisms and three presented one Gram-positive and one Gram-negative agent. One of the patients with two Gram-positive organisms was also po sitive for HSV-1 by qPCR analysis. From these five patients, four were contact lens wearers and one had a history of several previous intrao cular surgeries. All had ulcers >2 mm.

Four eyes harbored fungi and one eye presented Acanthamoeba growth in cultures. Although these eyes were clinically diagnosed as being potential bacterial corneal ulcers, microbiology studies were able to reveal the causative microorganisms and guide appropriate management. None of those patients were positive for herpes viruses by qPCR analysis.

Ten of 65 eyes (15.3%) were positive for viruses in the qPCR analysis. Of those, VZV was identified in one eye of one patient (a 93-year-old female), and HSV-1 was identified in nine eyes. All had ulcers >2 mm. Ocular comorbidities or severe systemic diseases were present in 7 of the 65 eyes (70%) (Table 3).

The mean age of patients who had positive qPCR results for herpes was higher than that of those who were negative for herpes: 60.4 and 42.2 years, respectively (p=0.01) (Table 4).

The control group comprised 25 patients with clinical diagnoses of herpes epithelial keratitis. All had dendritic lesions at presentation. Twenty-one of these patients were positive for HSV-1 by qPCR analysis (84%).

DISCUSSION

HSV-1 and VZV were identified in this study using qPCR of samples of infectious keratitis with bacterial growth in cultures, indicating a potential co-infection. Systemic and ocular comorbidities were found more frequently (70%) in cases that were positive for both virus and bacteria than in cases that were positive only for bacteria (32%). Neo vascular glaucoma with several previous intraocular surgeries, systemic immunosuppression due to HIV infection, chronic steroid use, severe allergy, and dry eye disease were found in 7 of 10 patients with a confirmed co-infection, and 4 of the 10 patients (40%) from this group were older than 70 years. This is a large proportion compared with the general study population, in which 9 of the 65 patients (14%) were older than 70 years. Statistical analysis revealed that the mean age of patients with positive qPCR results was significantly higher than that of patients with negative results (60.4 and 42.18 years, res pectively; p=0.01, Table 3). In addition, there was a tendency to a higher prevalence of positive PCR results in patients older than 70 years (p=0.07). Therefore, advanced age could be a factor for herpes infection.

A possible study limitation was the detection of latent HSV-1 and -2, as well as VZV. However, the target of the real-time primers and probes were amplified genes (the beta-globin gene in HSV-1 and -2, and the gene for the ORF29 protein in VZV), which are expressed almost ex clusively during active viral replication. HSV-1 latency in the nuclei of neurons of regional ganglia is well described^(2,3). As such, when clinical samples are collected from lesions located in the periphery of an axon, in the conjunctiva or mucosal surfaces, the virus is actively replicating and not latent. Besides, there are limited genes expressed during latency, and in designing the primers for this study, sequences that are expressed only during the active phases of infection and not during latency were chosen. Therefore, it is fairly certain that replicating HSV that was causing the disease was detected rather than latent viruses^(15-17).

The control group, which was composed of eyes with only active dendritic epithelial keratitis, showed a high positivity in qPCR analysis (84%). This strengthens the hypothesis of positive PCR results during active viral replication. Most of the eyes from the control group that had negative cultures and qPCR results had ulcers <2mm. The small amount of material collected could be responsible for the negative results from both culture and qPCR.

CONCLUSION

Microbiology studies are essential in infectious keratitis because different etiologic agents can manifest with similar clinical profiles. Molecular techniques show that herpes virus may be present in pa tients with bacterial corneal ulcers and qPCR may be useful in its detection and treatment, so as to avoid steroids, or appropriately admi nister antiviral drugs to patients who are positive for herpes by PCR. Cost-effective studies should be carried out to evaluate the impact of such molecular techniques.