Borrelia burgdorferi sensu lato can be cultivated from various tissues and body fluids in modified Kelly-Pettenkofer medium (MKP), Barbour-Stoenner-Kelly II (BSK-II) medium, and commercially available BSK-H medium [3, 18]. The yield of Borrelia culture from clinical samples is usually low with a higher rate of positivity from skin biopsy compared to blood and CSF cultures [18]. Borrelia cultivation and isolation is a time-consuming and demanding procedure. Successful cultivation can take up to 12 weeks [18]. In addition, preparation of culture media is demanding and expensive. Nonetheless, cultivation is the golden standard to prove the presence of active infection by Borrelia. Slow growth along with low clinical sensitivity [17, 19] do not make it a diagnostic tool for Lyme disease, and for this specific reason it will not be deeply considered in the present article.

Microscopic detection of Borrelia burgdorferi sensu lato has not been frequently used in clinical practice for confirmation of Lyme disease, mostly due to the sparseness of organisms in clinical samples [17]. Warthin-Starry stain, which consists of silver impregnation of microorganisms, was the oldest method used to search for Borrelia in histologic preparations in the 1990s, but recently only two publications have reported on the use of this stain for Borrelia detection [20, 21]. Among microscopic techniques focus floating microscopy (FFM), a modified immunohistochemical technique that combines several strategies to detect minuscule organisms in tissue sections, appears to be the most promising one for direct Borrelia detection, because it can be successfully applied on fresh material, nitrogen-frozen material, and paraffin-embedded tissues, in many cases on blocks older than 30 years [22]. Focus-floating microscopy has been recently proposed as the possible golden standard for detecting cutaneous borreliosis [22]. Not only a polyclonal antibody to Borrelia, a vivid chromogen (3-amino-9-ethylcarbazole) and the purposeful omission of counterstaining are necessary, but the technique also requires the histologist's patience [23]. By meticulously examining the entire slide horizontally and vertically, Eisendle et al. [22] have been able to find organisms with 98% sensitivity in classic examples of cutaneous borreliosis [22]. Since its first appearance in 2007 [22], FFM has been applied to several studies [21, 24-35], enabling detection of Borrelia and allowing that a number of dermatological conditions, such as morphea and lichen sclerosus et atrophicus, could be attributed to Lyme Borreliosis [28]. Unfortunately, although FFM is promising and helpful to prove Borrelia infection in cutaneous atypical lesions, only a few centers have adopted this technique.

Direct detection of Borrelia in clinical specimens is typically accomplished by PCR. Advantages of PCR are sensitivity, simplicity and speed [36]. A huge amount of articles have been published reporting on the use of PCR assays for Borrelia detection in human specimens, but large comparative studies aimed to identify the best PCR method are still lacking [3]. PCR assays differ in terms of target (borrelia sequence targeted by primers), PCR approach (nested PCR, real time PCR, PCR followed by hybridization…) as well as type of samples (blood, biopsies, cerebrospinal fluid, synovial fluid…) [37, 38]. Different types of PCR protocols could be applied for Borrelia detection, namely nested PCR, PCR followed by hybridization, realtime PCR, the latter both in qualitative as well as quantitative setting [37, 38]. Several and different protocols, mostly in house, are in use in most laboratories. Differences in protocols also rely on the type of samples and extraction procedures together with the broad range of clinical manifestations of Lyme disease. Different methods could be used to extract DNA and they could have different yields. Moreover, DNA extracts could be contaminated by PCR inhibitory substances due to poor extraction procedures [37]. Because of Borrelia paucity in biological liquids and tissues the diagnostic sensitivity of PCR was found to be low when using low volumes of sample material [3, 37]. As reported by Nolte PCR assays to detect Borrelia from CSF, synovial fluids or urines need large volumes for extraction in order to increase the sensitivity of the method [37]. Low and transient spirochetemia and high spirochoetes’ tropism for tissues (joints, heart, meninges) may account for negative PCR results in blood and other biological fluids [38]. While PCR is highly sensitive for detection of B. burgdorferi sensu lato DNA in skin biopsy samples, such testing for patients with Lyme disease involving systems other than skin, has in general low sensitivity, with the exception of patients with Lyme arthritis [17]. PCR protocols for Borrelia detection have been developed by targeting plasmid as well as borrelial chromosome. The most frequent target for plasmid borrelial DNA is OspA which in some cases have also been used for Borrelia genotyping [36, 39-43]. Among chromosomal targets the most frequently assessed in PCR assays are flagellin [43-47] and 66 kDa protein [41, 44, 48, 49] which were historically the first targets analyzed by PCR and most recently the gene encoding the Borrelia 16S rRNA [36, 43, 48-50]. As some plasmids may be present in more than a copy per Borrelia cell a plasmid target based PCR could be more sensitive. The “target imbalance” in Borrelia assays could be one of the multiple issues affecting Borrelia detection by PCR [37].

Overall, standardization is the major concern with respect to PCR for Borrelia detection.

PCR detection of Borrelia DNA does not enable to distinguish between dead or living microorganisms, because it is a poor indicator of the viability of the bacteria [51]. To this regard, quantitative analysis of flagellin B mRNA of Borrelia has been shown to correlate well with the culture results in Lyme disease, thus demonstrating the bacterial viability [51]. Although RNA analyses for Borrelia could be a surrogate biomarker for Borrelia active infection it is laborious and not suitable for routine analyses [37].

Recently, several CE-IVD PCR assays have been developed for Borrelia detection, among them: the EliGene^®Borrelia kit (Elisabeth Pharmacon group, Czech Republic and UK); the Borrelia burgdorferi PCR kit (GeneProofa.s, Czech Republic), the real time PCR Borrelia burgdorferi sensu lato (Dynex, Czech Republic), the Borrelia burgdorferi s.l. kit (Biorn Diagnostics GmbH, Germany), the Borrelia Burgdorferi kit (Clonit S.r.l. , Italy), the BactoReal^®Borrelia Burgdorferi sensu lato (Ingenetix, Austria) and the Diarella Borrelia real time PCR Kit (Gerbion, Germany) [4-10]. Most of them are qualitative real time PCR tests, multiplexing both an internal host control (human housekeeping gene) and the specific Borrelia target. Detection of DNA of Borrelia species pathogenic to human (B. afzelii, B. garinii, B. burgdorferi sensu stricto) has been guaranteed by amplifying flagellin gene [5-7], but also OspA [9]. Those commercial kits could be used for blood, additionally tests could be applied to cerebrospinal fluid and urine [4-6, 9, 10], to ticks [4, 10] or ticks’ suspension [6]. Sensitivities of those tests are highly variable and not conventionally described. Overall those commercial assays could have an impact on the reproducibility of PCR assays for Borrelia detection only because they decrease the variability in the reagent preparations and primer design. To this regard, the use of commercial kits represents a sort of harmonization in the assay preparation rather than standardization. There are no scientific publications comparing the performance of those commercial products with respect to serology, cultivation and in house PCR protocols in Lyme disease patients and controls. Clinical usefulness and added value to routine of those commercial assays should be confirmed/assessed by large validation studies with internal and external quality controls.

Overall, the scenario of PCR analyses for Lyme disease confirmation is wide and having reliable, reproducible tool standardization is of paramount importance in all the phases of the PCR process, starting from sample selection (e.g., type of sample and the required volume to be processed) as well as the pre-analytical phases such as the extraction procedures.

Most commonly in ELISA, antigen mixtures composed of whole-cell sonicates of B. burgdorferi sensu lato are used as the source of antigen for the the individual or in combination detection of IgG, IgM, or IgA antibodies. Usually those sonicates are derived from strain B31 [2], a specific strain of Borrelia burgdorferi sensu stricto, which is the only organism established to cause Lyme disease in North America [60]. Since in Europe and Asia Lyme disease also results from infection by Borrelia garinii or Borrelia afzelii, as well as by the recently described Borrelia spielmanii in Europe; diagnostic tests using antigens from B. burgdorferi sensu stricto will not necessarily perform well for infections carried by other genospecies, although some do [60].

The Vmp-like sequence expressed (VlsE) protein is a surface-exposed lipoprotein encoded by a linear plasmid of B. burgdorferi B31 and it has been found to be highly immunogenic [17]. In the last decade, an ELISA using only a single synthetic peptide derived from the VlsE sequence (IR6 or C6 peptide) as the source of antigen has become commercially available and is quite often used in the first step of the serological analyses. Both the single C6 peptide and whole VlsE assays have been approved as alternatives to whole-cell ELISAs as first-tier tests [60]. In addition, the Immunetics C6 test (Immunetics Inc, MA, USA) has recently been evaluated as an assay that could be used in place of both steps of two-tiered testing, that is, as a simple ‘stand-alone’ test [60]. C6- and VlsE-based assays have the additional feature of detecting antibodies to Eurasian genospecies of Borrelia (i.e. B. gariniiand B. azfelii) as well as B. burgdorferi sensu stricto [60]. Moreover, in a multiplex bead based assay for the detection of serum antibodies to Borrelia burgdorferi sensu lato it has recently been shown that VlsE IgG had the highest diagnostic value even in patients with neuroborreliosis [61].

Regarding immunoblots, recommendations have been published by Centers for Disease Control and Prevention (CDC) on the number and types of IgM and IgG bands in order to consider immunoblot results as positive [59]. However, those rules could be applied in the US, but not to patients infected in Europe since immune response of European patients is restricted to a narrower spectrum of Borrelia proteins, compared with that shown by American patients [3].

Consequently, the existence of at least 3 pathogenic species requires species-specific interpretation rules [52].

Traditional blots are hard to standardize, as they are only semi-quantitative and reading them involves often subjective judgment about the significance of weak bands [60].