Heterophile antibodies are antibodies that agglutinate cells of other animal species that are mainly associated with mononucleosis due to EBV but may be, albeit rarely, also detected in other diseases[47]. They peak 2-5 wk after symptom onset and then decline rapidly, although they may rarely persist for 6-12 mo[35,48]. Between 85% and 90% of adolescents and adults are positive during the course of EBV infection: about 50% in the first week, and 60%-90% in the second and third[29]. However, only 50% of children aged 2-5 years are positive at any time during the course of infection, and only 10%-30% of those aged less than 2 years[2]. The rate of false negative results can therefore be high in young children, whereas false positive results are observed in only 2%-3% of the patients with autoimmune diseases[2].

Despite these limitations, detecting heterophile antibodies can be helpful in the case of primary infection mainly because of the simplicity of the tests. In routine clinical practice, these include agglutination tests using sheep, goat or horse red blood cells as antigens after absorption with guinea pig kidney extracts in order to remove natural antibodies against Forssman antigen (the Paul Bunnel reaction), and newer tests based on latex particles with adherent specific bovine antigen for heterophile antibodies (monospot assays). Recently, multiplex flow immunoassays (MFI) have been proposed because they are more sensitive in the case of acute infections[49]. However, given the level of false negative results, negative findings need to be followed by a search for specific antibodies[2,48].

The specific tests for anti-EBV antibodies use different substrates or antigens and various technologies, of which those commonly used for the routine screening of EA IgG, EBNA-1 IgG, VCA IgG and IgM are immunofluorescence assays (IFAs) or enzyme immunoassay (EIAs) with enzyme-linked immunosorbent assays (ELISA) and chemiluminescence immunoassay (CLIA) versions or newer multiplex flow immunoassay (MFI). IFAs usually use EBV-transformed B cell lines from patients with Burkitt’s lymphoma (e.g. the P3HR-1 or Raj cell lines)[2,23,45,50-52], whereas EIAs use purified native or recombinant proteins, synthetic peptides or fusion proteins (complete proteins or fragments of the proteins encoded by the EBV genes)[42,48]. The type and preparation of the antigens used are probably responsible for the differences in the sensitivity and specificity of the various assays[42,53,54]. IFA has been used as the reference method, although its sensitivity is the same as or less than that of EIAs[43], automated versions of which allow a large number of samples to be tested and are commonly used in laboratories with a large routine workload. The latest CLIAs using synthetic peptides with different cut-off values for VCA IgM and EBNA-1 IgG can better distinguish the stage of infection[55] and, in the case of samples that are simultaneously EBNA-1 IgG, VCA IgG and IgM positive, may help distinguish recent (transient phase) and past infection or reactivation[55,56].

In order to confirm the screening assays, various immunoblotting tests have been developed using viral lysates of EBV-transformed cells and recombinant antigens (line blots)[57]. It is considered that the latter is unaffected by antibodies against the cell material that can be found in patients with mononucleosis[23,58]. Some line blots use recombinant antigens coated on the solid phase, such as EBNA-1 (p72), VCA (p18 and p23), EA (p54 and p138) and MA (gp 350/250); the most recent also use IEA (ZEBRA). It has been reported that the sera of patients with acute infection show anti-p23 IgG, anti-p55 IgG and anti-p138 IgG, but not anti-p72 IgG, whereas the sera from patients with past infection show anti-p23 IgG, anti-p72 IgG and anti-p18 IgG[23]. Kinetic studies have found that a strong IgG response to p72 is not observed until 20 d after disease onset. As anti-p72 IgG and anti-p18 IgG are present in patients with past infection, but not in those with acute infection[23], anti-p18 IgG can be considered equivalent to EBNA-1 IgG in terms of significance. In addition, as anti-p18 IgG is not lost in the case of immunosuppression[23], immunoblotting is especially useful in distinguishing acute and past infections in cases that are VCA IgG positive, but EBNA-1 IgG and VCA IgM negative.

Immunoblots for IgM have also been developed to detect VCA IgM in patients with acute infection, but not anti-p72 IgM[23], which may be useful for confirming the specificity of the VCA IgM detected by screening assays.

The IgG avidity test can assess the degree of IgG maturation. Avidity is low at the beginning of an acute infection, but increases when the immune response matures[59-61]. For example, the maturation kinetics of VCA IgG lasts several weeks and in some cases up to 3 mo after symptom onset[62,63]. Avidity can be measured using an EIA, IFA or immunoblotting[23,57,63]. Two aliquots of the same sample are tested in parallel for the presence of IgG antibodies: one is not treated, and the other is treated with substances that dissociate the antibodies from the antigens (usually 8 M urea). Since the dissociation depends on antibody avidity, the ratio between the treated and untreated sample defines the degree of avidity. A search for avidity can therefore be used to estimate the duration of a primary infection, and differentiate acute and past infection[23,57,63-65].

The published data depend on the examined antibody and therefore the type of antigen used for the test. In addition to a mix of antigens, it is possible to use the whole or specific VCA (p23 or p18), IEA (ZEBRA), EA (p138 or p54) or EBNA (p72). Avidity varies depending on the kinetics of the various antibodies[63,66], e.g. in VCA IgG, it has been reported that avidity is low in the samples collected during the first 12 wk after symptom onset, thus indicating recent infection[67]. With the passing of time, the avidity of VCA IgG may become borderline or high when the avidity of EA IgG is still low[63]. Immunoblotting, which has various antigens coated on the solid phase, can be used to visualize the avidity of various antibodies simultaneously (Table 4), although the maturation of anti-IEA and EA antibodies (BZLF1, p138 and p54) does not seem to indicate immune status as antibodies with a low degree of avidity can be found even during reactivation. Consequently, the manufacturer points out that a reduction in the intensity of the EA and IEA bands alone is not an index of recent infection[68].

The limitations of avidity testing are the individual maturation rates of antibodies and the fact that the tests cannot be used in newborns because of the presence of maternal antibodies.

A number of different methods, techniques and protocols have been used to determine the presence of EBV DNA and measure viral load[69-72]. Dot blotting, Southern blotting, PCR and in situ hybridization have all been applied to various materials, but their differences in sensitivity and specificity have led to the results that need to be considered cautiously[28] as they vary from laboratory to laboratory[73,74]. More recent studies indicate that real-time PCR is particularly sensitive[28], and very useful for defining infection status, especially in immunocompromised patients[45,75,76] and those at risk of developing EBV-related disorders[45]. However, there is still no consensus concerning the best material to use, units of measurement, or the quantitative levels requiring intervention or predicting prognosis[16,74,77-79]. This means that particular care is necessary when comparing the data of different studies[73]: for example, the units of measurement include copies per milliliter, copies per microgram of DNA, copies per 100 000 leukocytes, and copies per positive cell[77]. The targets used may also vary from one method to another: LMP2, BKRF1 or BamHI-W (EBNA-1), BNRF1 (membrane protein), BXLF1 (thymidine kinase), BZLF1 (ZEBRA), BALF5 (viral DNA polymerase) or BHRF-1 (transmembrane protein).

Furthermore, there is much debate concerning the material that should be used to search for EBV DNA, such as whole blood, peripheral blood mononuclear cells (PBMCs), plasma or serum[13,80]. There is also the problem that incorrectly stored whole blood can cause EBV DNA to leave the intracellular compartment and give rise to false positive results in plasma or serum, and false negative results may be due to nucleases that are capable of partially degrading plasma EBV DNA[81].

In general, the best material used to search for EBV DNA depends on where it is, and varies during the course of the disease[13]. The virions produced during primary infection spread in peripheral blood[82,83], and it is also possible to determine the EBV-free or fragmented DNA coming from apoptotic cells[83], and the B cells transformed during the latent phase also pass into the bloodstream. EBV DNA can therefore be determined in serum or plasma as well as in PBMCs[84]. In patients with primary infection, it is frequently detected in whole blood (PBMCs and plasma/serum) within 14 d of symptom onset[85-89]. After the initiation of an immune response, viral load decreases slowly in PBMCs, but rapidly in plasma/serum, and it becomes undetectable after 3-4 wk[90-92], whereas memory cells with EBV may remain latent for a long time in blood. However, it must be kept in mind that there may be individual variations due to individual differences in kinetics, and viral load may increase after an initial decline, and in some cases, it may take as long as a year or more before it reaches stably low levels. Finally, even when this level is reached, the blood of a healthy carrier contains 1-50 copies of EBV DNA per million white blood cells, whereas EBV-DNA is almost always undetectable in plasma or serum[82,85,93-96]. The presence of plasma/serum EBV-DNA is therefore considered a sign of primary infection[13] or reactivation, and the viral load correlates with disease severity[85,88,92]. A search for EBV DNA may be more sensitive than serology in the early stages of the disease[89], and some studies have found that it correlates better with clinical acute infection than the avidity of VCA IgG[86]. However, in immunocompetent patients with acute infection, it is not usually necessary to look for EBV DNA as serology is sufficient except in cases with negative or doubtful serological findings in which there is a strong clinical suspicion of infection[89,97,98].

A search for EBV DNA is particularly important in immunocompromised patients with an incomplete humoral response and patients who have received transfusions or immunoglobulins that confound serological test results[28]. It has been reported that immunocompromised patients have higher baseline viral levels than healthy carriers[99,100], which decline after treatment. A search for EBV DNA is also useful in patients with EBV-related tumors, except for those with AIDS-related brain tumors in whom the blood levels are low because of the blood-brain barrier[99,101]. In EBV-related cancers, episomal or naked EBV DNA from apoptotic tumor cells is found in serum and plasma[84,102], which may also contain tumor cells with latent EBV infection[13] and virions from a small number of tumor cells undergoing lytic infection. The most suitable material varies with the tumor[13] and depends on where EBV DNA is mainly found during the course of the disease (PBMCs or serum/plasma).

VCA IgM usually appears at the same time as VCA IgG, but because they can be detected earlier, a profile of isolated VCA IgM is usually thought to indicate an early stage of acute infection. Nevertheless, it is recommended to assess the specificity of the result because it may be made aspecific by interfering rheumatoid factor and autoantibodies, or cross-reacting factors such as HCMV or parvovirus B19[67]. A search for these factors and immunoblotting for IgM may be as helpful as the determination of other parameters of acute infection such as heterophilic antibodies or HBV DNA[98]. A search for anti-EA (D) IgG can also be useful as about 85% of the patients with acute infection are positive for these antibodies for up to 3 mo after the onset of symptoms[23,24].

In some cases, VCA IgM may not be produced or appear 1-2 wk after VCA IgG, or present for a very short time or at low concentrations and it may not be detected by conventional tests[2,32,103,104]. In addition, about 5% of the patients do not produce EBNA-1 IgG after EBV infection[103,105] or their levels remain below the limit of detection[37,105]. Furthermore, even when they are produced, they may be lost over time particularly, but not exclusively, in immunocompromised patients[23,103,106]. Consequently, isolated VCA IgG may be found in cases of past infection with the loss or disappearance of EBNA-1 IgG, as well as in cases of acute infection with the delayed or early disappearance of VCA IgM. This pattern can be found in approximately 7% of cases in routine laboratory practice and in about 8% of all subjects with at least one marker of EBV infection; it also tends to become more frequent with advancing age[107].

Such cases require further diagnostic investigations (Table 5), including immunoblotting for IgG, avidity tests for VCA IgG, searches for viral genome, or heterophile antibodies or EA (D) IgG[43], or a repetition of the test after about 30 d in order to identify any change in the antibody profile. However, this last option inevitably delays the diagnosis until the second sample is collected, and physicians tend to avoid it if the symptoms improve over time, especially in the case of children, they may find it traumatic, which means the second sampling usually involves only a small number of patients[107].

Immunoblottting for IgG allows the interpretation of cases with isolated VCA IgG[23]. Some tests use recombinant p72 antigen (EBNA-1), p18 and p23 (VCA), p54 and p138 (EA) and gp 350/250 (MA). As anti-p18 IgG can be considered equivalent in meaning to EBNA-1 IgG (because they are both produced late during EBV infection), their presence indicates past infection, and their absence indicates an acute infection. Immunoblotting can therefore distinguish cases with the loss or non-appearance of EBNA-1 IgG, which however develop an anti-p18 IgG response, from those with primary EBNA-1 IgG negativity (acute infection)[23,108]. Using this method to examine cases of isolated VCA IgG, about 20% was identified as acute infections and 80% as past infections[107].

A VCA IgG avidity test can also be used to distinguish acute and past infection[64,67], which are respectively characterized by low and high levels of avidity; studies of patients with isolated VCA IgG have found low levels in about half of the cases, and high levels in the remaining half[75,86]. When using avidity test, it is important to have all the patients’ temporal clinical data in order to interpret the results more appropriately.

The detection of EBV DNA is particularly useful in the case of serologically indeterminate EBV infection[75,86,98] because its presence in serum or plasma indicates primary infection or reactivation[13]. Some studies have found that the presence of EBV DNA in patients with isolated VCA IgG is usually associated with a low degree VCA IgG avidity[75,86] (especially when the sample is taken early in the course of suspected infectious mononucleosis), and in some cases it has been found to correlate better than VCA IgG avidity[86]. Furthermore, in cases of past infection diagnosed on the basis of a high level of VCA IgG avidity, none of the serum samples were positive for EBV DNA[86].

A search for heterophile antibodies may be useful in patients with isolated VCA IgG if the result is positive, although there have been reports of their presence for 6-12 mo[28,35,48]. About 60% of patients with acute infection and isolated VCA IgG present heterophile antibodies, as against none of those with past infections[107]. However, nothing can be said in the absence of heterophile antibodies, especially in children aged less than 5 years, and other tests must be used.

One possible alternative is to search for EA (D) IgG because about 85% of patients with acute infection are positive for EA (D) IgG for up to 3 mo after symptom onset[23,24]. However, as high titers can be found during EBV reactivation[26], as well as in 20%-30% of healthy subjects with past infection[2,25] and in patients with nasopharyngeal carcinoma[27], the diagnostic use of these antibodies is still being debated. In general, a search for EA (D) IgG alone does not identify the stage of disease[43,109], but combining these antibodies with other parameters can be useful in laboratory diagnoses[110] because it has been found that only 12% of the patients with isolated VCA IgG and past infection are EA (D) IgG positive, as against 90% of subjects with isolated VCA IgG and acute infection[111].

VCA IgM may persist for several months after an acute infection[35], and may also reappear during EBV reactivation[33,112]. Consequently, EBNA-1 IgG, and VCA IgG and IgM may be simultaneously present in patients with primary EBV infection if VCA IgM persist and EBNA-1 IgG have already been produced (a phase that has been variously defined as “recent infection”, “primary infection, transient phase or convalescence”, “past infection, IgM persisting”), or in those with reactivation and the simultaneous presence of VCA IgM and EBNA-1 IgG[42]. Reactivation is still relatively rare and often short in immunocompetent subjects, and is generally considered of no clinical relevance[113,114]; however, it can cause serious complications in immunocompromised patients.

This serological pattern is uncommon (approximately 5% in normal routine laboratory practice)[107], and further diagnostic tests are needed to distinguish transient infection and reactivation (Table 6). First of all, it is important to verify the specificity of VCA IgM because there may be false positive results[115-117] during the course of infection with other pathogens, such as HCMV, parvovirus B19, Toxoplasma gondii, hepatitis A or HIV[118-124]. It has been shown that primary HCMV infection often causes a further antibody response in anti-EBV IgM[125], and cross-reactive antibody responses against conserved epitopes are well known among herpes viruses, such as the glycine-alanine epitope shared by EBV and HCMV[119]. In some studies, up to about 50% of VCA IgM in patients with past infection may be due to non-specific reactivity[67] and a positive result should be confirmed (e.g. by immunoblotting for EBV IgM)[124]. It is also possible to test for cross-reactivity or polyclonal IgM, especially in the case of HCMV and parvovirus B19[98], and a search for HCMV or parvovirus B19 DNA can also help in some cases[98]. Moreover, as false positive reactions may also result from the presence of autoantibodies or rheumatoid factor, it can be useful to look for these interfering factors. However, it needs to be pointed out that cold agglutinins, rheumatoid factor and autoantibodies can be found for a short period during the course of EBV mononucleosis because of the polyclonal activation of B-cells[126], therefore, the presence of rheumatoid factor does not automatically mean a falsely positive VCA IgM result. Finally, EBNA-1 IgG positivity in patients with primary infection may also result from aspecific reactivity, which can be detected by immunoblotting for IgG antibodies[124]. Furthermore, the recently developed EIA for antibodies to EBNA-1 IgG based on recombinant or synthetic peptides may be more sensitive than its predecessors, and allow their identification early in the course of primary EBV infection[124].

Once the specificity of the results obtained has been established, additional diagnostic approaches are necessary (Table 6). In addition to repeating the test after a reasonable period of time in order to detect any changes in the antibody profile, VCA IgG avidity has proved to be particularly useful because low levels of avidity have been found in the course of recent infection, and high levels in the course of past infection or reactivation[62-64]. Some studies have found that fewer than half of immunocompetent patients with this profile and a low degree of avidity have primary infection, and about 20% or less a reactivation that is probably clinically insignificant[112,124]. In order to fully understand the IgG avidity test and evaluate the results in relation to antibody maturation time, it is important to know how long after the onset of symptoms the blood sample was taken.

The use of heterophile antibodies is controversial because some studies have shown that they are very sensitive (94% of infected patients with transient infection)[124], whereas others have found very few cases of simultaneous positivity for EBNA-1 IgG, VCA IgG, VCA IgM and heterophile antibodies[109].

The use of molecular biology seems to be a rather delicate question. Patients with latent infection have an almost constant number of circulating infected B cells in peripheral blood and, in the case of reactivation, these differentiate into plasma cells, leading to the start of the replicative cycle and increased EBV DNA levels in PBMCs and serum/plasma. It has been reported that a search for EBV DNA in PBMCs and serum/plasma is important for an immediate diagnosis of reactivation but, although this is true in the patients who are followed up over time in order to detect any changes in viral load, the finding of EBV DNA in a single sample should not necessarily be seen as a sign of reactivation[95]. Consequently, as EBV DNA is present in cases of reactivation or primary infection, it is unlikely that testing one sample will be able to distinguish the two situations[75,95].

Persistent or reactivated EBV infection is characterized by high titers of EA (D) IgG, especially in immunocompromised patients[26,29]. After reactivation, the levels of these antibodies increase with VCA IgG, whereas the levels of EBNA-1 IgG decrease. An increase in the titer of EA (D) IgG can therefore be considered a useful marker of reactivation. However, as EA (D) IgG can also be found in 85% of primary infections and 20%-30% of past infections, and simultaneous VCA IgM, VCA IgG, EBNA-1 IgG and EA (D) IgG positivity has been seen in both the transient phase and reactivation[76], the detection of anti-EA (D) IgG seems to be useful only if serial sampling is possible. Other antibodies have been found in cases of reactivation. It has been suggested that EBNA IgM may be useful in identifying reactivation[114]. The combination of negative EBNA IgG and positive EBNA IgM, in addition to a high degree of avidity for VCA IgG within 3 mo of an acute infection, should reliably indicate reactivation rather than primary EBV infection[67].

It has also been suggested that VCA IgA may be a marker of reactivation. These antibodies reach a peak level 3-4 wk after primary infection and decline slowly, but may last indefinitely. Different studies have found them in 35%-74% of acute cases[34,127,128], but they were also seen in 10% of healthy subjects. High levels were also found in patients with immunodeficiencies, recurrent parotitis, multiple sclerosis or nasopharyngeal cancer, as well as in pregnant women and elderly subjects[26,129-132]. Consequently, a search for VCA IgA is considered useful only in the diagnosis and management of patients with nasopharyngeal carcinoma[133-136].

EA (R) IgG, EA IgM, EA IgA, VCA IgG-3 and the EBNA-1/EBNA-2 ratio have also been used as markers of reactivation in some studies[18,95,114,137-139] but, as many of these antibodies are also present in primary infections, and some even in past infections, the usefulness of a single sample in patients with the simultaneous presence of EBNA-1 IgG, VCA IgG and IgM has yet to be evaluated.

Finally, the use of different cut-off values ​​proposed by the most recent CLIAs for determining VCA IgG, VCA IgM and IgG EBNA-1 in parallel seems to be a promising approach as it should make it possible to distinguish the different stages of EBV infection, especially the difference between transient infection and past infection or reactivation[55,56].

VCA IgM usually appears at the same time as VCA IgG antibodies, but they disappear completely within a few weeks[32-37], whereas patients are VCA IgG positive for the rest of their lives[32]. EBNA-1 IgG cannot usually be detected during the first 3-4 wk after the onset of clinical symptoms[32,140] and as VCA IgG antibodies persist but EBNA-1 IgG antibodies may disappear, especially in the case of immunosuppression[23,103,106], a pattern of isolated VCA IgG without VCA IgM or EBNA-1 IgG has been widely documented as mentioned above, whereas the presence of EBNA-1 IgG without VCA IgG is generally considered implausible[43]. However, some immunoblotting studies have found that 2% of subjects who have experienced previous infections may be negative for VCA IgG (anti-p23 or anti-p18)[23] and, although rare, this pattern can be found in routine laboratory tests and may represent about 1.7% of all EBNA-1 IgG-positive samples[39]. An isolated EBNA-1 IgG pattern therefore gives rise to some interpretative doubts: it is possible to envisage aspecific EBNA-1 IgG, a lack of VCA IgG production or their loss over time. At clinical level, the problem is to discover whether the patient has experienced a past infection or is still susceptible. As the doubt is to distinguish past infection and non-specificity, it is useless to search for heterophile antibodies or EBV DNA. The presence of EA (D) IgG may help indicate a past infection as approximately 20%-30% of healthy subjects with a history of EBV infection have EA (D) IgG[2,23,25], but nothing can be said in the case of negativity. However, immunoblotting for IgG can solve the problem as it can confirm the real presence of EBNA-1 IgG. One immunoblotting study found that such patients were not only all anti-p72 (EBNA-1) positive, but also anti-p23 (VCA) positive, which was not detected by the screening tests using p18 antigen alone to detect VCA IgG[140]. It therefore seems that EBNA-1 IgG does not exist without the presence of at least one VCA IgG, such as anti-p23.