EFNS-ENS guidelines for the use of PCR technology for the diagnosis of CNS infections

The EFNS-ENS consensus recommendations for the usage of PCR technology for the diagnosis of infections of the nervous system have been published. A summary of the guideline is given below.

Polymerase chain reaction (PCR) is a simple and rapid, easy- to- use approach to diagnose infections by amplifying nucleic acids. The reliability of PCR technology for the diagnosis of neurological infections is dependant on the type of pathogen

The PCR involves the in vitro enzymatic synthesis of millions of copies of a specific DNA segment. 
1. The reaction is based on the annealing and extension of two oligonucleotide primers that flank the target region in duplex DNA
2. After denaturation of the DNA, each primer hybridizes to one of the two separated strands such that extension from each 3′ hydroxyl end is directed toward the other. 
3. The annealed primers are then extended on the template strand with a DNA polymerase. 

These three steps (denaturation, primer binding, and DNA synthesis) represent a single PCR cycle. Consequently, repeated cycles of denaturation, primer annealing, and primer extension result in the exponential accumulation of a discrete fragment whose termini are defined by the 5′ ends of the primers. 

The length of the products generated during the PCR is equal to the sum of the lengths of the two primers plus the distance in the target DNA between the primers. PCR can amplify double or single-stranded DNA, and with the reverse transcription of RNA into a cDNA copy, RNA can also serve as a target.

Quantitative PCR, uses precision optics and DNA-binding fluorescent dyes or fluorescent labels to monitor amplification in real-time.

Viruses

PCR not only allows a specific diagnosis in an individual patient but can also define the spectrum of disease caused by a particular virus and when applied to large numbers of clinical samples can help understand the epidemiology of neurological infections. 

Several viruses can be looked for in the same CSF or other sample using the technique of multiplex PCR in which several pairs of primers specific for particular viral sequences are used. Where more than one virus is detected in a CSF sample, the significance of the virus detected has to be carefully evaluated, especially if such a dual viral infection is made more likely by immunosuppression as occurs during Human Immunodeficiency virus (HIV) infection 

While evaluating PCR results the sensitivity and specificity of the particular assay are important factors to take into consideration ie. the possibility of false negative and false positive results. The timing of the CSF sample and the physical conditions such as specimen storage can be important

Using quantitative PCR real-time PCR allows the determination of the viral load in a patients' blood or CSF. Although this is not carried out routinely in most laboratories, it has been used in some cases to assess the severity of the viral disease burden and/or the prognosis, examples being Cytomegalovirus (CMV) and JC virus infections

Besides CSF Viral PCR can also be carried out on other tissues such as peripheral blood, brain, or other tissue biopsy specimens.

Herpes simplex virus

1. The sensitivity of CSF PCR in HSE is 96% and the specificity is 99% . Hence CSF PCR can be recommended as a highly reliable method of diagnosing HSE without the need for brain biopsy (Level A).
2. It is important to carry out CSF PCR for both HSV-1 and HSV-2 since HSV-2 is also associated with mild or atypical of cases of HSE, particularly in immunosuppressed patients such as those with HIV infection as well as neonatal HSE caused by HSV-2.
3. Timing of PCR: The CSF PCR for HSE can be negative during the early stage of the infection. Re-examination of the same negative CSF a few days later may yield a positive PCR result . The CSF PCR may also be negative if the sample is taken too late during the infection since the yield of virus is highest during the first week of the infection following which it falls .The guidelines recommend the optimum timing of CSF PCR at 2–3 days to 10 days after symptom onset (Level C). 
4. If HSE is strongly suspected even if CSF PCR obtained within the first 72 h of the onset of symptoms is negative it should be repeated few days later to obtain a definitive diagnosis (Level B).
5. Treatment with acyclovir does not reduce the chances of PCR detecting HSV during the first week of infection so such treatment should not influence the decision to carry out a PCR test (Level A). Once acyclovir has been started in a case of suspected HSE with a negative PCR, we recommend that it is continued for 14 days unless an alternative diagnosis has been established (Level C). 
6. Currently it is not recommended to repeat the CSF PCR in all patients after 14 days of acyclovir treatment. There is an on going US NIH anti-viral study group trial which also addresses this issue.
7. Quantitative PCR: Routine Quantitattive PCR to assess the viral load has not been shown to be a useful prognostic marker in an individual patient with HSE and hence is not recommended.

Varicella-Zoster virus

1. The sensitivity and specificity of VZV PCR in cases of VZV-associated neurological disease has been estimated at 80% and 98%, respectively.
2. CSF PCR for VZV should be considered in all cases of encephalitis of unknown cause . 
3. PCR is also helpful in proving the etiology in a spectrum of neurological conditions caused by VZV without reactivation including VZV vasculopathy, zoster sine herpete, and myelopathy, meningoencephalitis, and polyneuritis cranialis. It is important to think of this possibility in differential diagnosis. In such cases in addition to CSF PCR for VZV DNA other tests for eg detection of anti-VZV IgG in the CSF should also be carried out.
4. CSF anti-VZV IgG more sensitive than PCR in VZV vasculopathy

Cytomegalovirus

1. CSF PCR has a very high sensitivity and specificity for detecting CMV infection and is recommended in patients with suspected CMV associated neurological disease.
2. Quantitative CMV PCR may also be useful to correlate disease severity and monitor the efficacy of anti-viral therapy 

Epstein–Barr virus

1. EBV PCR on the CSF (Level C) is recommended in patients presenting with symptoms suggestive of the range of neurological conditions associated with EBV including encephalitis, aseptic neuritis, cerebellar ataxia, myelitis, and several peripheral nerve disorders including various types of acute radicultis, radiculoplexopathy, acute autonomic neuropathy, Guillain–Barre syndrome, and cranial neuropathies .
2. The sensitivity of CSF EBV PCR in AIDS patients with a suspected CNS lymphoma is very high, at 97% -100%
3. Quantitative PCR can also be used in such patients to predict the risk of developing non-Hodgkin CNS lymphoma and for monitoring the effects of chemotherapy 

Enteroviruses

1. Enteroviruses (EV) disease spectrum includes a non-specific febrile illness, aseptic meningitis, and encephalitis. A chronic meningoencephalitis may also occur in immunocompromised patients.
2. CSF Enterovirus PCR is highly sensitive and specific in patients suspected of having an EV infection and recommended for both accurate diagnosis and improved patient management.
3. Reverse transcription (RT)-PCR for EV provides a rapid and very accurate diagnosis of an infection in <24 h and much more quickly than is possible with standard viral culture (Class II). 
4. The CSF PCR for EV results should however be interpreted carefully in case of false negative results since some studies have shown that EV PCR of specimens from the respiratory and gastro-intestinal tracts yielded higher results than did CSF
5. Obtaining a specific diagnosis is particularly important since patients may be potentially treated with the antiviral agent pleconaril. 

JC Virus

1. JC Virus (JCV) is a polyoma virus that is the causative agent of progressive multifocal leukoencephalopathy (PML), a demyelinating infection of the CNS that occurs mainly in immunocompromised individuals, primarily those with AIDS and recently also associated with natalizumab therapy in multiple sclerosis patients.
2. The specificity of CSF PCR progressive multifocal leukoencephalopathy is excellent at 98.5–100% though the sensitivity is lower in the region of 50–82% (Class II).
3. CSF PCR to detect JCV DNA is now the established and routine method of diagnosis in PML and has superseded brain biopsy 
4. In cases where the PCR is negative in a brain biopsy should be seriously considered to obtain a definitive diagnosis. 
5. Quantitative PCR is also helpful since higher CSF JC virus loads have been found to be associated with shorter survival times and viceversa. JC virus loads have also been used to monitor the effects of antiviral therapy in PML patients

Human immunodeficiency virus

1. Diagnosis will already have been made on the blood
2. Quantitative PCR to measure the CSF viral load has been a valuable tool in assessing neurological involvement in HIV infection such as HIV-associated dementia and encephalitis and also to monitor therapy as the CSF viral load decreases markedly following highly active antiretroviral therapy (HAART).

Human T-cell lymphotropic virus-1

1. Human T-cell lymphotropic virus-1 (HTLV-1) is strongly associated with tropical spastic paraparesis and HTLV-1-associated myelopathy
2. Use of PCR is recommended (Level C) in the diagnosis of tropical spastic paraparesis and HTLV-1-associated myelopathy. 
3. The sensitivity and specificity of CSF PCR for tropical spastic paraparesis and HTLV-1-associated myelopathy has been reported as 75% and 98.5%, respectively. Combination of CSF PCR and anti-HTLV-1 antibody index useful in diagnosis (Class III Level C)

Bacteria

The PCR results are usually available well within 24–36 h for common bacterial infections and utilize low volume of CSF (≥1 ml) for analysis. 

PCR methods may employ one of the several available techniques:

1. Nested or semi-nested PCR with hybridization and sequencing, or use of universal primers and restriction endonuclease enzyme digestion. Nested approach is considered better in detecting meningeal infections with Borreila, Listeria, or Mycoplasma because of the low numbers of bacterial DNA and relatively few copies of 16S rRNA gene in the CSF
2. Probe-based real-time PCR is often preferred when using a multiplex PCR for simultaneous detection of different specific target sequences. Real-time quantitative multiplex PCR is considered a highly sensitive technique for fast identification of a causative pathogen of bacterial meningitis and can detect as few as two copies of Neisseria meningitidis, Streptococcus pneumoniae, and Escherichia coli, 16 copies ofListeria monocytogenes, and 28 copies of group B streptococcus whereas the sensitivity for broad-range 16S rRNA-based PCR was about 10–200 organisms per ml of CSF
3. Broad-range bacterial PCR is based on use of primers that recognize conserved regions of the genes encoding for eubacterial 16S ribosomal RNA (rRNA). The broad-range bacterial PCR combined with sequencing may be of particular advantage in rapid diagnosis and identification of the etiologic agent in community acquired bacterial meningitis .

Acute meningitis

1. Currently available PCR methods detect Hemophilus influenzae, N. meningitidis, S. pneumoniae, L. monocytogenes in CSF and have a sensitivity of 87–100% and specificity of 98–100%
2. Quantitative multiplex RT-PCR is commonly preferred for the detection of common pathogens of acute bacterial meningitis. 
3. PCR-based detection of bacterial pathogens is also considered more sensitive than culture in patients with ventricular catheters and suspicion of nosocomial meningitis or conventional cultures withprior antibiotic treatment
4. The positive predictive value of broad-range PCR is 98%, and the negative predictive value is 100%; ie a negative bacterial PCR assay virtually excludes the diagnosis of acute bacterial meningitis.
5. It is recommended that in house nucleic acid amplification methods for diagnosis of bacterial infections in CSF are deemed unreliable and should not be used in clinical practice (Class IV Grade C) owing to high inter-assay variability and low specificity
6. The advantage of uniplex vs multiplex quantitative RT-PCR is currently unclear (Class IV Grade C).
7. PCR-based diagnostic tools should be used only as an adjunct rather than a substitute for current methods of bacteriological diagnosis by conventional staining and culture.
8. In Future: it might be possible to rapidly diagnose bacterial infections by Gram stain-specific probe-based real-time PCR using 16S rRNA which will enable simultaneous detection and discrimination of clinically relevant Gram-positive and Gram-negative bacteria directly from blood samples 

Chronic meningitis

1. Tuberculous Meningitis:
   1. Quantitative RT-PCR has been shown to substantially increase diagnostic yield in TBM.
   2. A wide range of sensitivities have been reported for PCR-based tests forMycobacterium tuberculosis in CSF samples . 
   3. Using CSF filtrate than CSF sediment has been reported to give higher yield of positive result qualitative RT-PCR. 
   4. Since PCR-based methods are also prone to cross-contamination like conventional cultures, and the diagnostic specificity of PCR-based diagnosis of TBM may be compromised in endemic areas . 
   5. A negative CSF PCR result does not exclude the diagnosis of neurotuberculosis when CSF and imaging data suggests otherwise and one must be aware of false positives as well.
   6. In acute bacterial meningitis , CSF rapidly becomes sterile after antibiotic therapy and bacterial DNA may not be detectable beyond 8 h of treatment . However Mycobacterial DNA may persist for up to a month in CSF after starting therapy .Hence even if initial PCR result is negative, repeating quantitative RT-PCR test in successive CSF samples for M. tuberculosis may aid to diagnosis even if initial PCR result is negative
   7. The current consensus is that repeating CSF PCR within first 3 weeks may aid diagnosis in tuberculous meningitis if the initial result is negative (Class IV, Grade C).
2. Lyme neuroborreliosis: 
   1. In Lyme neuroborreliosis CSF PCR for Borrelia is probably useful as a diagnostic test only in very early stages and is not recommended as a diagnostic test for chronic Lyme disease or measure treatment response as a follow-up.
   2. However in patients with atypical forms of erythema migrans PCR on skin biopsies may become useful in the diagnosis of early Lyme borreliosis. 
   3. CSF- PCR is not presently a validated diagnostic test for Lyme neuroborreliosis (Class IV, Grade C).
   4. Direct microscopy and culture remain the gold standard of microbiological diagnosis of bacterial infections of central nervous system where feasible and current range of diagnostic bacterial PCR tests do not replace them (Class II Grade A).
   5. Commercially available and standardized quantitative RT-PCR is recommended for routine use in CSF samples (Class II, Grade A) of patients with suspected bacterial meningitis. 

Parasites

1. The diagnosis of a parasitic infection of the CNS rests upon clinical signs and symptoms, clinical history, travel history including geographic exposure and, finally, on laboratory techniques.
2. In resource poor areas light microscopy remains the diagnostic mainstay. 
3. Indirect methods like serology do not distinguish between past, latent, reactivated, or acute infection and is not useful in assessing response to therapy or prognosis. 
4. In immuno-compromised patients, indirect diagnostic methods (serology for antibody detection, etc.) have a very low sensitivity. 
5. Highly specific tests to detect antigen, for example, rapid antigen detection system (RDTS) luciferase immune precipitation system (LIPS), molecular-based approaches, in particular PCR , loop-mediated isothermal amplification (LAMP) , real-time (RT) PCR , and luminex technology have shown a high potential for use in diagnosing parasitic infestations with increased sensitivity and specificity .

Molecular-based diagnostic tests in protozoal infections of the CNS

Table below lists the protozoa that have the capacity to invade the CNS, causing neurological disease and details the molecular-based techniques used in their detection. Except cerebral toxoplasmosis the other molecular-based assays are still experimental diagnostic techniques and have not yet replaced serology or direct proof by light microscopy.

Molecular-based diagnostic tests in helminthic infestations of the central nervous system

1. Conventional PCR has been supplemented by nested and multiplex PCR as well as real-time PCR for the detection of several parasitic infestations and infections, respectively.
2. More advanced techniques as loop-mediated isothermal amplification (LAMP) and luminex-based assays have also been proposed as possible diagnostic techniques in parasitic diseases of the nervous system. 
3. Direct visualization or detection of the helminths, either the adult worm, the larval stage, or the eggs, be it in body fluids or biopsied material, still represents the golden standard of diagnosis

Fungi

The gold standard of diagnosis is positive cultures together with microscopy, antigen/antibody testing in serum, and CSF. However the slow growth of fungi in culture, cross-reactivity in case of antigen detection, and dependence on the demonstration of an antibody response or even by the failure to mount an adequate immune response are drawbacks. There is not enough data to recommend the routine use of PCR in fungal infections.

Histoplasmosis (Histoplasma capsulatum) :

1. Most common endemic mycosis in Europe
2. Chronic meningitis is the most frequent CNS manifestation of histoplamosis
3. Cerebral or spinal masses and encephalitis are less common. 
4. Fungal culture is the gold standard diagnostic test in non-CNS manifestations, but it may take up to 6 weeks 
5. CSF cultures usually do not yield growth.
6. Histoplasma capsulatum antigen and antibodies can be determined by different methods in CSF and are used to establish diagnosis. However, cross-reactivity with other dimorphic fungi and Cryptococcus species is seen in upto 50% of cases.
7. There is no commercial kit available for routine use for PCR in CNS histoplasmosis

Coccidioidomycosis (Coccidioides immitis)

1. Chronic basal meningitis is the most common coccidioidal CNS manifestation.
2. Occasionally, CNS coccidioidomycosis presents as meningoencephalitis or intracranial mass lesion.
3. Coccidioides sp. grows in culture within 2–5 days 
4. Coccidioides sp. is isolated from CSF only in a third of patients with CNS manifestations 
5. CSF antibodies can be detected in up to 70% of patients with meningitis during the initial analysis, and in the majority of cases later. 
6. Currently, there are no standardized CSF antigen detection methods
7. No PCR is available for commercial use.

Blastomycosis (Blastomyces dermatitidis)

1. CNS manifestations of blastomycosis are rare, mostly cranial/spinal epidural abscess, meningitis, and brain abscess in AIDS patients.
2. Definitive diagnosis of blastomycosis requires growth of the organisms in culture. 
3. CSF culture is rarely positive, and only stereotactic brain biopsy may be diagnostic.
4. cytological examination of CSF may sometimes reveal Yeast forms.
5. CNS blastomycosis may be diagnosed by presence of antibodies in serum and CSF, cross-reactivity with other fungi is possible. 
6. Chemiluminescent DNA probes have been developed but not standardised or commercially available.

Cryptococcosis (Cryptococcus neoformans/gattii)

1. Chronic basal meningitis is the most frequent CNS manifestation of cryptococcal disease causing subacute dementia or visual symptoms.
2. Culture alone is generally not the method of choice. 
3. The diagnostic mainstay is antigen detection with >90% sensitivity and specificity. This test in CSF can be positive early in infection. 
4. Microscopic examination of CSF using India ink stain is diagnostic in up to 80% of AIDS patients and about 50% in non-immunocompromized.
5. The usage of CSF PCR for the diagnosis of suspected CNS cryptococcosis is likely to be of value

Candidiasis (Candida albicans and other C. species)

1. Candida species are the fourth leading cause of fungal bloodstream infections and associated with a mortality rate of up to 50%.
2. CNS Candidiasis can develop in the setting of disseminated candidiasis with microabscesses of the brain parenchyma or as candida meningitis in association with a foreign body (e.g. catheter, ventricular shunt) or other CNS invasive procedures (e.g. surgery).
3. Blood cultures are considered the gold standard for routine diagnosis of invasive candidiasis but are time-consuming.
4. CSF, stains and routine cultures have a low yield in identifying the pathogen in chronic Candida meningitis. 
5. Serological diagnosis of CNS candida infection, either by antigen detection (mannan) or by antibody determination, has not been validated. 
6. PCR protocols for detection of candidal infection have been described and require probes for different subspecies. PCR may also assess mutations associated with resistance to antifungal medication.

Aspergillosis (Aspergillus fumigates and other species)

1. There are numerous species of aspergillosis recognized, but most cases of CNS infection are attributed to A. fumigates, A. flavus,A. terreus, and A. versicolor. Major CNS manifestations include hemorrhagic infarction, abscess, and meningitis; less frequent are mycotic cerebral aneurysm and granuloma.
2. Culture is insensitive and diagnosis requires non-culture-based methods. 
3. Two antigen assays, the [1,3]-beta-D-Glucan assay (sensitivity 87%), a relatively non-specific assay and the galactomannan assay (>95% sensitivity and specificity), an Aspergillus-specific antigen, are commercially available and used in clinical routine for blood specimen and sometimes in CSF samples.
4. A combination of the galactomannan (antigen) assay and PCR may improve diagnosis.
5. The sensitivity of PCR testing for Aspergillus is limited during antifungal treatment.
6. The usage of CSF PCR for the diagnosis of suspected CNS aspergillosis is likely to be of value

Mucormycosis/zygomycosis (Mucor, Rhizpous, Rhizomucor, Absidia, Cunninghamella, Apophysomyces, Saksenaea)

1. Mucormycosis is an acute and aggressive fungal infection, which can develop as isolated cerebral mucormycosis (16%), extension to the brain from rhinocerebral mucormycosis (69%) or via the hematogenous route (15%).
2. Diagnostic techniques include histology, culture, and PCR. The diagnosis is mostly made by a combination of histology and culture.

Conclusion

1. The main use of PCR technology is to the diagnosis of infections caused by viruses followed by bacterial infections of the CNS with the notable exception of tuberculous meningitis.
2. The efficacy of PCR for the diagnosis of both protozoal infections and helminthic infestations has also been established but far from becoming routine in resource-poor countries where such infections are prevalent.
3. There are class IV evidence studies reporting the feasibility of CSF PCR for evaluating CNS manifestations by Histoplasma, Coccidioides, and Candida, and of tissue for CNS mucormycosis. There is not enough evidence at present to recommend the use of PCR as a routine diagnostic tool in these cases.

Original citation:

Steiner, I., Schmutzhard, E., Sellner, J., Chaudhuri, A. and Kennedy, P. G. E. (2012), EFNS-ENS guidelines for the use of PCR technology for the diagnosis of infections of the nervous system. European Journal of Neurology, 19: 1278–1291. doi: 10.1111/j.1468-1331.2012.03808.x [ Abstract]

Other resources:

CME Quiz on this article