Uses and Misuses of the Polymerase Chain Reaction in the Diagnosis of Ophthalmic Disease

One of the most exciting advances in ophthalmic research in recent years has been the development of polymerase chain reaction (PCR) as a research tool to link pathogens to particular diseases. This powerful molecular biological technique enables the replication of a single deoxyribonucleic acid (DNA) molecule to 10 million or more nearly identical copies in the space of several hours, facilitating the analysis of infinitesimal quantities of nucleic acids (Science. 1988;239:487-491). When DNA belonging to a pathogen is amplified in this way, the technique can provide evidence of infection and, thereby, assist in diagnosis. The eye is particularly well-suited to this method of diagnosis, because the aqueous and vitreous humors (the normal sites of biopsy) are relatively acellular, and this allows for a high signal to noise ratio for PCR with minimal interference from host DNA. Additionally, the differential diagnoses for pathogens underlying many forms of ocular inflammatory disease are relatively limited (Surv Ophthalmol. 2001;46:248-258). This article will help clinicians to understand the uses and limitations of this powerful technique.

PCR Enhances Disease Understanding

Fuchs’ heterochromic cyclitis is an uncommon, unilateral uveitis that typically features a white and quiet eye coupled with chronic cell, cataract, and elevated intraocular pressure (IOP). Quentin and colleagues studied aqueous fluid from patients with Fuchs’ heterochromic cyclitis and found elevated antibody titers to rubella in the aqueous humor of these patients (Am J Ophthalmol. 2004;138:46-54). However, in a subset of patients the researchers were able to directly demonstrate rubella nucleic acid (RNA) by reverse-transcription PCR. The combination of intraocular antibody production and direct detection of pathogen DNA makes a very compelling case for linking this pathogen to this unusual condition.

Similarly, recent work has demonstrated cytomegalovirus (CMV) DNA in the aqueous humor of 5 patients with recalcitrant unilateral anterior uveitis accompanied by elevated IOP. In this study, all patients were treated with antiviral medications for CMV and experienced substantial improvement. Since CMV is also found latent in white blood cells, one must be careful in interpreting this finding. However, de Schryver and associates demonstrated very high intraocular antibody titers to CMV (and not to other herpes family viruses), strongly suggesting a linkage in immunocompetent patients between CMV and this form of uveitis (Br J Ophthalmol. 2006;90:852-855).

Meanwhile, Hariprasad et al recently demonstrated the etiology of the choroiditis in an immunocompromised patient who was admitted for cellulitis. Peripheral blood smears suggested the presence of an algae (Prototheca wickerhammi). Although the patient did not survive his infection, the remaining family members allowed analysis of the eyes at autopsy. The findings demonstrated positive PCR reactivity for the algae from the choroid, confirming this linkage (Arch Ophthalmol. 2005;123:1138-1141). This is the first time algae has been identified as a cause of choroiditis in humans.

Requirements for PCR

The process of performing PCR on an aqueous or vitreous biopsy begins with purification of the DNA sample. Highly cellular samples such as solid tissue biopsies or massively infiltrated vitreous are typically subjected to stringent DNA extraction techniques similar to those used in forensics. Less cellular specimens can be processed directly, although endogenous inhibitors of PCR must be inactivated prior to performing the reaction (J Clin Microbiol. 1995;33:2643-2646). A typical way of doing this is to boil DNA samples 15 minutes before performing PCR.

The reaction mix for PCR consists of the purified DNA sample, appropriate buffers, nucleotide triphosphates (the building blocks for DNA), a thermostable DNA polymerase, and most importantly oligonucleotide primers specific for the pathogen(s) being tested. This is an important point to remember when performing PCR: the reaction can only produce amplification products from the primers used in the reaction. In other words, if one fails to account for a particular pathogen on differential diagnosis, one will never find it with PCR.

The reaction itself is a thermal cycle. A machine alters the temperature between one permissive for polymerase activity, one permissive for oligonucleotide binding, and one that dissociates the nascent DNA strands. Typically 35 to 45 cycles are performed, which takes approximately 2 hours for most primer sets. The products may then be analyzed via agarose gel electrophoresis. It is essential that negative and positive controls be run at the same time as the patient sample to evaluate for possible false negative and false positive results. The latter are particularly problematic.

Indications for PCR

PCR is generally performed under 3 circumstances: when the diagnosis is unclear, during an atypical clinical course, and/or concomitant with another procedure. The most common diagnostic dilemma is one in which media opacity prevents evaluation of the fundus. The dense vitritis seen in both acute retinal necrosis syndrome (ARN), and ocular toxoplasmosis can create this difficulty, as can cataract or corneal opacity. Atypical presentations will often be suspected when a case fails to follow the anticipated course—for example, a case presumed to be ARN syndrome which does not respond to antiviral therapy. Finally, when another procedure requires the obtaining of a DNA sample, this often provides an opportunity for PCR diagnostics. For example, when foscarnet is injected into the vitreous of patients presenting with ARN syndrome, this also allows sampling of the vitreous for PCR diagnostics to determine if the ARN syndrome is due to herpes simplex or varicella zoster. Such a distinction has prognostic relevance, as herpes simplex-associated ARN has a very high rate of associated encephalitis (Am J Ophthalmol. 2000;129:166-172).

Currently, most diagnostic PCR is performed for posterior segment inflammatory disease. However, many large hospital laboratories now routinely perform PCR reactions for herpes simplex, varicella zoster, cytomegalovirus, and Toxoplasma gondii. It is worthwhile to determine if the laboratory being used has been certified according to Clinical Laboratory Improvement Amendment (CLIA) standards, because this assures that appropriate quality-control measures are in place. It is also important to coordinate carefully with the laboratory prior to submitting the patient sample, since different laboratories have different preferences for how the sample is obtained and transported. For example, laboratories are often used to performing PCR diagnostics on cerebral spinal fluid samples, which are typically several milliliters. However, PCR can be performed on samples as small as a few microliters, and the laboratory may have to alter its procedures to accommodate such small samples.

If the laboratory is in the same building as the operating room, the sample can be drawn into a sterile 1 cc syringe, the needle removed, and the syringe capped and placed on ice. One hundred ul of aqueous and ~0.5 cc of vitreous per sample is typically removed in this manner. If the sample needs to be sent to an outside laboratory, it is typically frozen on dry ice and shipped by overnight courier. It is important that the sample not be subjected to freeze-thaw cycles, and it is essential that a full clinical history be provided to the laboratory analyzing specimen. The clinician should also specifically request which organism DNA should be amplified.

Pitfalls of PCR

The advantages of PCR—its exquisite sensitivity and specificity—are also the sources of potential pitfalls. With regard to the technique’s high sensitivity, because even single genomes of pathogens can potentially be amplified, false positives are possible. These may occur from laboratory contamination or from amplification of commensal DNA in the sample. For example, it is common for individuals to shed herpes simplex virus in tears, and this virus can be picked up during a biopsy and result in a false positive result. One means for circumventing this difficulty is to use quantitative PCR (qPCR).

In qPCR, rates of amplification are measured using a fluorescent dye; appropriate calibration allows for precise quantification of the starting material. This is the technology used commonly to calculate viral loads in patients with human immunodeficiency virus (HIV). Using this technique, it is possible to distinguish a commence organism, which will have a relatively low copy number, from that of an active infection, which will have a high copy number (Arch Ophthalmol. 2002;120: 1534-1539).

A second pitfall involving PCR is its very high specificity. Even a single base pair mismatch between oligonucleotide primer and pathogen can result in loss of the amplification product. Therefore, false negative results are possible with organisms that may mutate highly or have many strains. It is advisable to test several different primer sets before concluding that a negative PCR result indicates the absence of a pathogen.

It is important to remember that PCR can only detect organisms for which a primer set has been included. Margolis and colleagues recently tallied their results following PCR for vitritis where clinical suspicion for any particular entity was low. They found there was very little utility to performing PCR in settings where the differential diagnosis is not well defined (Am J Ophthalmol. 2006;141:584-585).

Future Uses of PCR

One of the more intriguing new uses of PCR is in the diagnosis of bacteria and fungi. These microbes have highly conserved ribosomal genes (16S for bacteria, and 18S and 28S for fungi), and these genes are highly amplified in a microbial genome. By using primers to recognize these conserved sequences, it is possible to amplify for all bacteria or fungi in a sample. This technique has been previously applied to culture negative endophthalmitis. Okhravi and colleagues performed PCR on 22 vitreous aspirates, 7 of which were culture negative. In this study, all culture-positive aspirates had PCR results matching culture, and all culture-negative aspirates were positive for bacterial DNA via PCR (Invest Ophthalmol Vis Sci. 2000;41:3474-3479). Several were novel organisms, which had not been previously cultured. One technical drawback to using 16S PCR has been the need to clone and sequence the amplification products in order to identify the underlying organisms. However, techniques including denaturing gradient gel electrophoresis obviate the need for sequencing and allow rapid typing of microbial organisms (Appl Environ Microbiol. 1993;59:695-700).

High throughput DNA sequencing has also advanced remarkably in the past several years. It is now possible with massively parallel, high throughput sequencing to obtain sequences from tens of millions of base pairs per day (Nature. 2005;437:376-380). This opens up the possibility of sequencing all the DNA in a pathology specimen in order to determine what is derived from a pathogen in what is derived from the host. Such an approach would circumvent the current difficulty of having to identify specific primers prior to PCR amplification. It is likely that this technology will become clinically available in the next several years.

References

Author Disclosure

The author states that he has no financial relationship with the manufacturer or provider of any product or service discussed in this article or with the manufacturer or provider of any competing product or service.