Molecular techniques are changing the face of diagnostics, from identifying secondary tumour origins and gene expression levels right through to pre-natal and pre-implantation genetic testing. Such techniques provide a much deeper understanding of disease processes, particularly cancer, allowing accurate differential diagnosis, bespoke surveillance and treatment strategies and also identification of reliable prognostic indicators. Gynaecological cytology, however, is comparatively less established in molecular diagnostics than other disciplines.
Cervical cancer is the eleventh most common cancer in females, with less than 3000 cases diagnosed each year and since the introduction of the NHS cervical screening programme in 1988 the number of deaths from cervical cancer has reduced considerably. Presently, there are over 200 genotypes of Human Papilloma Virus (HPV) identified, inclusive of at least 15-18 specific high-risk HPV (HR-HPV) types that are well known causal agents of almost 99% of cervical cancer cases. The Pap test, however, is far from a perfect strategy with around a third of pre-cancerous and/or cancerous cases being misdiagnosed as normal, together with rare detection of early HPV lesions due to the cytology screener relying heavily on the pathognomonic presence of the koilocyte perinuclear halo (see opposite), a morphological change that not every HPV infection will present with.
For equivocal Pap results, where women present with atypical squamous cells of undetermined significance (ASC-US), a fateful decision for patient follow-up has to be made between referral to colposcopy or a six-month recall for another Pap test. This is problematic because ASC-US is potentially considered at high risk of progression to squamous intraepithelial lesions due to approximately 50% of these samples containing HR-HPV infection; but also many cases of ASC-US are transient, which is particularly common in women under 30 and so it is crucial to ascertain the correct triage to avoid over-treatment. The Pap test, therefore, can effectively mask conditions signalling the need for additional colposcopic investigations and more aggressive treatments. This subsequently reduces the sensitivity and negative predictive value (NPV) of the Pap test, thus increasing the possibility of cancer progression during the recall interval. With such an unfavourable limitation and the highly subjective nature of this diagnosis, cervical screening methods are turning to adjunctive molecular techniques, which can detect HPV infections and assess risk levels at a much earlier stage.
The status of the HPV infection is necessary to identify patients that are at a greater risk of cancer progression (also associated with persistent HR-HPV infection). HR-HPV testing, therefore, has an important clinical role in the triage of women with ASC-US Pap results by only referring HR-HPV positive patients to colposcopy. Presently in the UK the implementation of molecular HPV detection assays remain largely restricted to the private and research sectors and here are a few of them:
Hybrid Capture II
The most widely used test is the nucleic acid hybridisation assay, Hybrid Capture II (HC-II). This is carried out using the same liquid based cytology (LBC) sample as collected for the Pap preparation, to detect the presence or absence of DNA from any of the following 13 HR-HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68. Using a cocktail of probes, the hybrids of the HR-HPV types and complimentary HPV RNA probes are detected using chemiluminescence and the light intensity is measured as relative light units (RLUs). The HC-II test for HR-HPV infection in ASC-US samples instantly increases the sensitivity as a cervical cancer screening technique by providing a distinct marker for suspicious samples requiring colposcopy referral.
Shortly after the HC-II was introduced, concerns were raised about the probe cocktail and the possibility it produced a high degree of cross-reactivity with untargeted HPV types, including some low-risk HPV (LR-HPV) genotypes; HC-II also had major problems with reproducibility, particularly in HR-HPV weakly-positive cases. The purpose of the HPV test is to assess the level of risk of cancer progression; therefore, cross-reactivity with LR-HPV has a detrimental impact on the accuracy of the test; possibly resulting in over-treatment of the patient, which could lead to future problems such as miscarriage.
Polymerase Chain Reaction
To eliminate cross-reactivity with untargeted HPV types, polymerase chain reaction (PCR) techniques using LBC and biotinylated primers (such as AMPLICOR, PGMY, MY11/09), GP5+/6+ and SPF10 HPV), to amplify short sequences within the L1 region in the HPV genome (see below) and indicate the presence of HR-HPV with product absorbancy levels from a colourimetric reaction. PCR techniques require half the volume of sample needed for HC-II, they have a similar sensitivity but a higher specificity than HC-II, however, ease of contamination in the PCR methods increase the possible risk of false positives in this case, rather than the cross-reactivity seen in HC-II, thus potentially lowering the PPV.
With relatively low PPV for both HC-II and PCR methods, an alternative molecular technique can be used, known as the Cervista HPV assay, which uses “Invader Technology” to exploit the occurrence of single base pair alterations that determines specific HR-HPV genotypes. Two simultaneous isothermic reactions involving a fluorescence resonance energy transfer (FRET) probe, results in the emission of a fluorescent target-specific signal. There are two Cervista assays; the Cervista HPV HR, which identifies the same pool of HR-HPV types as HC-II with the exception of an additional HPV-66 target and the Cervista HPV 16/18, which identifies the two most prevalent HR-HPV types, HPV-16 and HPV-18, found in 50-65% and 7-20% of cervical cancer cases, respectively. The Cervista HPV 16/18 assay is an improved technique over HC-II and PCR because it reduces cross-reactivity and false positives, thus increasing the PPV; it also limits HR-HPV positivity to just HPV-16 or HPV-18, as opposed to producing a simple generic “High-Risk” positivity consisting of multiple possible types. Cervista HPV testing also boasts a substantial reduction in technician-operating time, compared to the HC-II, which needs regular technician interaction.
Given that HPV-16 and HPV-18 are the two most prevalent genotypes in cervical lesions, this must not diminish the significance of the oncogenic potential of other HR-HPV genotypes and so detection of specific HPV genotypes would prove a very effective diagnostic tool for assessing risk of cancer progression. The two most significant applications for HPV genotyping are for the identification of a persistent infection, which can denote a greater risk of cancer progression if HR-HPV is detected and also categorisation of HPV-16 and HPV-18 positive women for immediate referral to colposcopy and those with other lower risk HR-HPV genotypes, to be allocated less aggressive treatments. HPV genotyping would subsequently enable effective and bespoke patient monitoring and treatment strategies, but as an HR-HPV positive result with HC-II and PCR does not differentiate between HR-HPV types and Cervista HPV16/18 does not differentiate between HPV-16 and HPV-18, these assays are unable to provide such definitive results.
A similarly sensitive assay, but one that can distinguish between infections of HPV-16 and HPV-18 and infections from a pool of other HR-HPV types in a single reaction is the Roche Cobas 4800 HPV test, which demonstrates no cross-linking with untargeted HPV types, thus improving PPV of specific HPV-type detection. Using Real-Time PCR (RT-PCR) to quantify the HR-HPV present in the sample, four specific probes are used, one for HR-HPV 16, another for HR-HPV 18, a generic probe for the remaining 12 HR-HPV types and an internal control probe to assess the quality of the DNA in the sample. Automated RT-PCR techniques have good reproducibility and they increase the sensitivity of the HR-HPV test.
Beyond identifying the presence of the two most common HR-HPV infections, HPV PCR products from LBC samples can be used for specific genotyping for the identification of other significant HR-HPV such as HPV-31 and HPV-33, which are also associated with high risk of high-grade cervical lesion progression. Two examples of genotyping techniques are the digene HPV Genotyping Reverse Hybridisation (RH) Test and the Roche Linear Assay (LA), or Reverse Line Blot (RLB). The RH test allows the identification of 18 HR-HPV types (HPV 16, 18, 26, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 68, 73 and 82) by using specific complementary probes, immobilised on strips of nitrocellulose (see below), whereas the LA/RLB assay can identify 37 HPV types, both high- and low-risk (HPV 6, 11, 16, 18, 26, 31, 33, 34, 35, 39, 40, 42, 43, 44, 45, 51, 52, 53, 54, 55, 56, 57, 58, 59, 61, 66, 68, 70, 71-CP8061, 72, 73, 81-CP8304, 82-IS39, 82-MM4, 83-MM7, 84- CP6108 and 84-MM8). RH and LA/RLB are particularly useful in the diagnosis of both persistent and mixed HPV infections, with LA/RLB also able to identify LR-HPV; they both demonstrate high sensitivity and specificity, together with rapid, high-throughput and good reproducibility.
HPV testing could not only reduce patient’s anxiety through reassurance of the test result, it will also increase the screening recall interval and reduce the NHS expenditures and the time spent unnecessarily assessing transient ASC-US. Recent proposals have suggested that HPV testing should be a mandatory adjunct to primary cervical cytology or even be the sole primary screening test, with cervical cytology triage reserved only for HPV positive tests. In reference to the latter, it must be considered that cervical cytology relies heavily on the quality of the sample being taken and LBC does not reduce the occurrence of ASC-US. The resulting low sensitivity and high subjectivity of the Pap smear diagnosis, therefore, calls for a more sensitive and reproducible assay for this triage.
Fluorescence in-situ Hybridisation
One approach focuses on the expression of a cell cycle inhibitor, P16INK4A, as a possible marker representing the link between HR-HPV and cervical neoplastic progression. A study into P16INK4A expression by fluorescence in situ hybridisation (FISH) on cytology LBC samples, using probes for the 9p21 (P16 INK4A) locus and a centromeric satellite probe for chromosome 9, highlighted some interesting results. The P16 INK4A copy number positively correlated with CIN grade and HPV viral load, therefore, effectively identifying more cases with, or at risk of progressing to, high-grade lesions. FISH is a quantitative, non-subjective approach – could this be a more promising triage technique?
Fortunately, the costs of molecular tests are decreasing and not only do they reduce the number of unnecessary referrals, they also reduce the time ineffectively spent by medical staff on such referrals, which in turn benefits the workload throughput and the economy. It is, however, the accuracy and reproducibility of the assay that is paramount if HPV testing is being considered as a potential primary adjunct, or even a sole primary screening test.
With current NHS legislations only suggesting HPV testing viability in equivocal Pap results, perhaps an improved approach would be a routine RH test for HR-HPV genotyping, as a primary adjunct with cytology, to identify HPV-16 and HPV-18 infections requiring immediate colposcopy referral and all other HR-HPV infections to undergo further p16INK4A FISH tests to assess their risk of progression to high-grade lesions and to measure the urgency for colposcopy. From a clinical perspective, the significance of LR-HPV testing at this time may be difficult to prove given the slow implementation and acceptance by health services of HR-HPV testing.
Molecular techniques for testing of HPV, despite being in the early stages of implementation, are demonstrating some important comparative advantages to cervical cytology alone. They provide a more rapid throughput, good reproducibility, are convenient due to their utilisation of the LBC sample, they provide excellent quality assurance and have the potential to remove the requirement for serial cytology. Dependent on the assay, when compared to the Pap test, the majority show extended diagnostic value due to increased sensitivity and specificity together with providing more information about the sample in both routine and follow-up assessments by presenting definitive indications of higher risk patients requiring prioritised referral for further treatment.
What must be considered before implementing HPV testing are the possible implications of providing HPV results as part of a routine screening programme; such as when compared with Pap preparations, HPV testing can reassure women with ASC-US, that a negative HPV result represents a relatively low risk of the atypical cells progressing to cancer. HR-HPV infection in women presenting with normal Pap preparations, however, would not least be confusing to the patient, but also be quite distressing and so the value of such molecular testing is questionable because it would seem irrational to employ aggressive treatment strategies in the absence of cervical lesions and/or other abnormal symptoms such as irregular bleeding or pain, despite the molecular results.
The distinct shortgae of widely available HPV tests in the NHS is a possible indicator of the difficulties faced in adopting a molecular technique worthy of the gold standard recognition and until the professionals of the cervical screening programme are convinced of the necessity, let alone the accuracy of a molecular test, all these advances in HPV testing and cervical cancer screening will, no doubt, remain redundant.
Koiloctyes: adapted from http://www.flickriver.com
PCR primer targets: adapted from http://www.moleculaire-pathologie.nl/moleculairedag270209/Snijders.pdf
Reverse Hybridisation test: adapted from Geraets, D.T., Heideman, D.A., de Koning, M.N., Snijders, P.J., Meijer, C.J., van Doorn, L.J., Quint, W.G., (2009). High genotyping concordance between the digene HPV Genotyping RH Test and the Reverse Line Blot genotyping assay on GP5+/6+-PCR products. Journal of Clinical Virology : The Official Publication of the Pan American Society for Clinical Virology. 46 Suppl 3 S16-20.