Introduction

Cervical cancer prevention programmes aim to screen the population, detect and treat high-grade cervical intraepithelial neoplasia (CIN) 2 and 3+ and follow up patients. For over 50 years, Pap cytology followed by colposcopy was the cornerstone for detection of CIN. Persistent infection with oncogenic human papillomavirus (HPV) has been established as the necessary cause of invasive cervical cancer. High-risk HPV (hrHPV) deoxyribonucleic acid (DNA) have been identified in 99.7% of specimens of invasive cervical cancer and in over 95% of the immediate cervical cancer precursors, namely high-grade squamous intraepithelial lesions (HSILs), such as CIN 3 or carcinoma in situ . Detection of hrHPV DNA is now recommended for use as a method of primary screening of cervical cancer, as a marker for risk of progression to cervical cancer in the case of borderline smears, and as a follow-up method to detect recurrence after treatment of CIN2 and 3+. The commercial availability of several HPV tests has widened the scope of cervical cancer screening and prevention using this method.

Methods of human papillomavirus testing

Screening assays based on high-risk HPV DNA represent a group of qualitative or semi-quantitative multiplex assays in which the DNA of the targeted HPV types is detected using mixtures of probes (probe cocktails) for several HPV types with similar clinical characteristics.

Molecular techniques can be broadly divided into those technologies that are not amplified, such as nucleic acid probe tests, and those that use amplification, such as polymerase chain reaction (PCR). Amplification techniques can be further divided into three separate categories: (1) target amplification, in which the assay amplifies the target nucleic acids (e.g. PCR); (2) signal amplification, in which the signal generated from each probe is increased by a compound-probe or branched-probe technology; and (3) probe amplification, in which the probe molecule itself is amplified (e.g. ligase chain reaction).

Hybrid Capture technology uses specific ribonucleic acid (RNA) probes, hybridisation, antibody capture, and signal amplification to allow rapid, standardised testing of genetic material. The Digene Hybrid Capture 2 ^® (HC2) hrHPV DNA test (Qiagen Gaithersberg Inc., MD, USA) identifies 13 high-risk HPV types (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68) and five low-risk types (6, 11, 42, 43 and 44) using two different RNA probes: probe B for high-risk types and probe A for low-risk types, complementary to the genomic sequence of the respective HPV types. In 2000, the US Food and Drug Administration (FDA) approved HC2 for triage of women with atypical squamous cells of undetermined significance (ASCUS) Pap smear results to determine the need for referral to colposcopy. In 2003, it was approved for use in conjunction with the Pap smear in women aged 30 years or older to assess the absence or presence of high-risk HPV types. The HC2 test is the most commonly used HPV test, and is presently widely regarded as the reference standard for HPV testing.

The HC2 test is relatively easy to carry out and is commercially available in a 96-well microplate format, with in-built positive and negative controls. Denaturation of HPV DNA allows incubation of the single-stranded HPV DNA with the RNA probes, whereupon the HPV DNA is hybridised with each of the probe cocktails, resulting in the formation of specific HPV DNA–RNA hybrids, which are then captured by antibodies bound to the wells of a microtitre plate that recognise specific DNA–RNA hybrids. Excess antibodies and non-hybridised probes are removed, and the immobilised hybrids detected by a series of reactions that give rise to a luminescent product that can be measured by a luminometer. The intensity of emitted light, expressed as relative light units (RLUs), is proportional to the amount of target DNA in the specimen, providing a semi-quantitative measure of the viral load. The recommended cut-off value for a positive test is 1.0 relative light unit (which is equivalent to 1 pg HPV DNA/mL sampling buffer, corresponding to 5000 genomes per test well, a clinically significant threshold). The result gives a positive result when the DNA of any one of the types is present above a certain threshold and does not provide information on the specific types of HPV detected.

HC2’s high-risk probe cocktail, however, may cross-react with HPV types that are not represented in the probe mix, such as types 53, 66, 67 and 73. Cross-reaction with types not represented in the probe cocktail could lead to false–positive results, with a consequent decrease in test specificity unless the types are definitely associated with cervical cancer, in which case their detection would improve the sensitivity of the test.

In 2009, the FDA approved Cervista ^® HPV HR (Hologic, Inc., Bedford, MA, USA) for HPV testing for the same indications as HC2. Cervista ^® tests for 14 hrHPV types, the 13 types identified by HC2 as well as type 66 and Cervista ^® HPV 16/18 test that individually identifies these two most carcinogenic hrHPV types.

Amplicor ^® HPV (Roche Molecular Systems, Alameda, CA, USA) is a test based on PCR for the same 13 hrHPV types as HC2, approved for use in Europe, Canada and Japan. Similarly to HC2 and Cervista ^® , Amplicor ^® expresses the results of the tested group of hrHPV types as positive or negative. Amplicor ^® is based on standard PCR amplification and detection of PCR products on microwell plates. Briefly, after sample preparation, a 165 bp long part of the HPV L1 gene and fragments of the human beta-globin gene are co-amplified with a mixture of biotin-labelled primers. Aliquots of denaturated amplicons are added to separate wells of microwell plates coated with either hrHPV probes or beta-globin-specific oligonucleotide probes. After a washing procedure, bound hybrids are detected with a biotin avidin-horseradish peroxidase assay. The original Amplicor ^® test was set up with a cut-off value for positivity of 0.2.

Cobas 4800 ^® HPV (Roche Molecular Systems Inc., Alameda, CA, USA) reports pooled results for 12 hrHPV types and for HPV 16 and 18. It was only recently approved by the FDA for the same indications as HC2 and Cervista ^® . The Linear Array ^® HPV genotyping test (Roche Molecular Systems Inc.) is a PCR-based test using PGMY 09/11 primers in which 37 HPV types (low risk and high risk) are hybridised on a single probe. It is considered as a reference standard for HPV genotyping and has been used in many trials.

New studies have been published on clinically useful RNA based assays. The most relevant transcripts for diagnostic purposes are those encoding viral oncoproteins E6 and E7. The detection of viral messenger RNA (mRNA) can be carried out by reverse transcriptase PCR or by nucleic acid sequence-based amplification. For the latter, three commercially available assays that detect E6 and E7 transcripts are currently available: PreTect ^® HPV-Proofer, based on nucleic acid sequence-based amplification technology, NucliSENS EasyQ ^® HPV, based on original HPV-Proofer assay, and APTIMA ^® HPV Assay (Gen-Probe, San Diego, CA, USA), which is a transcription-mediated amplification-based assay, which allows the detection of E6/E7 mRNA transcripts of 14 HPV types.

RealTime High Risk HPV ^® test (Abbott Molecular, Des Plaines, IL, USA) is a real-time PCR assay based on concurrent individual genotyping for HPV 16 and HPV 18 and pooled detection of 12 other HPV types (31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68). Amplification of human beta globin is used as an internal control. The assay uses four channels for the detection of fluorescent probes: one for the detection of the internal control (human beta globin); a second one for the detection of HPV 16; a third one for the detection of HPV 18; and a fourth one for the detection of the remaining 12 hrHPV types. The assay turnaround time is 6–8 h for 96 samples, depending on the method used for DNA extraction.

Methods of human papillomavirus testing

Screening assays based on high-risk HPV DNA represent a group of qualitative or semi-quantitative multiplex assays in which the DNA of the targeted HPV types is detected using mixtures of probes (probe cocktails) for several HPV types with similar clinical characteristics.

Molecular techniques can be broadly divided into those technologies that are not amplified, such as nucleic acid probe tests, and those that use amplification, such as polymerase chain reaction (PCR). Amplification techniques can be further divided into three separate categories: (1) target amplification, in which the assay amplifies the target nucleic acids (e.g. PCR); (2) signal amplification, in which the signal generated from each probe is increased by a compound-probe or branched-probe technology; and (3) probe amplification, in which the probe molecule itself is amplified (e.g. ligase chain reaction).

Hybrid Capture technology uses specific ribonucleic acid (RNA) probes, hybridisation, antibody capture, and signal amplification to allow rapid, standardised testing of genetic material. The Digene Hybrid Capture 2 ^® (HC2) hrHPV DNA test (Qiagen Gaithersberg Inc., MD, USA) identifies 13 high-risk HPV types (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68) and five low-risk types (6, 11, 42, 43 and 44) using two different RNA probes: probe B for high-risk types and probe A for low-risk types, complementary to the genomic sequence of the respective HPV types. In 2000, the US Food and Drug Administration (FDA) approved HC2 for triage of women with atypical squamous cells of undetermined significance (ASCUS) Pap smear results to determine the need for referral to colposcopy. In 2003, it was approved for use in conjunction with the Pap smear in women aged 30 years or older to assess the absence or presence of high-risk HPV types. The HC2 test is the most commonly used HPV test, and is presently widely regarded as the reference standard for HPV testing.

The HC2 test is relatively easy to carry out and is commercially available in a 96-well microplate format, with in-built positive and negative controls. Denaturation of HPV DNA allows incubation of the single-stranded HPV DNA with the RNA probes, whereupon the HPV DNA is hybridised with each of the probe cocktails, resulting in the formation of specific HPV DNA–RNA hybrids, which are then captured by antibodies bound to the wells of a microtitre plate that recognise specific DNA–RNA hybrids. Excess antibodies and non-hybridised probes are removed, and the immobilised hybrids detected by a series of reactions that give rise to a luminescent product that can be measured by a luminometer. The intensity of emitted light, expressed as relative light units (RLUs), is proportional to the amount of target DNA in the specimen, providing a semi-quantitative measure of the viral load. The recommended cut-off value for a positive test is 1.0 relative light unit (which is equivalent to 1 pg HPV DNA/mL sampling buffer, corresponding to 5000 genomes per test well, a clinically significant threshold). The result gives a positive result when the DNA of any one of the types is present above a certain threshold and does not provide information on the specific types of HPV detected.

HC2’s high-risk probe cocktail, however, may cross-react with HPV types that are not represented in the probe mix, such as types 53, 66, 67 and 73. Cross-reaction with types not represented in the probe cocktail could lead to false–positive results, with a consequent decrease in test specificity unless the types are definitely associated with cervical cancer, in which case their detection would improve the sensitivity of the test.

In 2009, the FDA approved Cervista ^® HPV HR (Hologic, Inc., Bedford, MA, USA) for HPV testing for the same indications as HC2. Cervista ^® tests for 14 hrHPV types, the 13 types identified by HC2 as well as type 66 and Cervista ^® HPV 16/18 test that individually identifies these two most carcinogenic hrHPV types.

Amplicor ^® HPV (Roche Molecular Systems, Alameda, CA, USA) is a test based on PCR for the same 13 hrHPV types as HC2, approved for use in Europe, Canada and Japan. Similarly to HC2 and Cervista ^® , Amplicor ^® expresses the results of the tested group of hrHPV types as positive or negative. Amplicor ^® is based on standard PCR amplification and detection of PCR products on microwell plates. Briefly, after sample preparation, a 165 bp long part of the HPV L1 gene and fragments of the human beta-globin gene are co-amplified with a mixture of biotin-labelled primers. Aliquots of denaturated amplicons are added to separate wells of microwell plates coated with either hrHPV probes or beta-globin-specific oligonucleotide probes. After a washing procedure, bound hybrids are detected with a biotin avidin-horseradish peroxidase assay. The original Amplicor ^® test was set up with a cut-off value for positivity of 0.2.

Cobas 4800 ^® HPV (Roche Molecular Systems Inc., Alameda, CA, USA) reports pooled results for 12 hrHPV types and for HPV 16 and 18. It was only recently approved by the FDA for the same indications as HC2 and Cervista ^® . The Linear Array ^® HPV genotyping test (Roche Molecular Systems Inc.) is a PCR-based test using PGMY 09/11 primers in which 37 HPV types (low risk and high risk) are hybridised on a single probe. It is considered as a reference standard for HPV genotyping and has been used in many trials.

New studies have been published on clinically useful RNA based assays. The most relevant transcripts for diagnostic purposes are those encoding viral oncoproteins E6 and E7. The detection of viral messenger RNA (mRNA) can be carried out by reverse transcriptase PCR or by nucleic acid sequence-based amplification. For the latter, three commercially available assays that detect E6 and E7 transcripts are currently available: PreTect ^® HPV-Proofer, based on nucleic acid sequence-based amplification technology, NucliSENS EasyQ ^® HPV, based on original HPV-Proofer assay, and APTIMA ^® HPV Assay (Gen-Probe, San Diego, CA, USA), which is a transcription-mediated amplification-based assay, which allows the detection of E6/E7 mRNA transcripts of 14 HPV types.

RealTime High Risk HPV ^® test (Abbott Molecular, Des Plaines, IL, USA) is a real-time PCR assay based on concurrent individual genotyping for HPV 16 and HPV 18 and pooled detection of 12 other HPV types (31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68). Amplification of human beta globin is used as an internal control. The assay uses four channels for the detection of fluorescent probes: one for the detection of the internal control (human beta globin); a second one for the detection of HPV 16; a third one for the detection of HPV 18; and a fourth one for the detection of the remaining 12 hrHPV types. The assay turnaround time is 6–8 h for 96 samples, depending on the method used for DNA extraction.

Human papillomavirus deoxyribonucleic acid testing for primary screening

Cervical cancer screening based on exfoliative cytology has been effective in reducing incidence and mortality rates of cervical cancer, especially squamous cell carcinoma, when coverage exceeds 70–80% of the population at a frequency of one to five annually. In the USA, most cervical cancers are now seen among women who do not have regular screening. Sasieni et al., however, reported that 47% of women in the UK who developed stage 1B1 or worse invasive cervical cancer before the age of 70 years had an adequate previous screening history. Cytology has certain limitations. First, results are dependent on a high-quality sample being collected during examination. Second, the test requires the identification of morphological changes within cells and interpretation is highly subjective. Last, this method of screening is particularly repetitive, which can lead to a greater number of interpretive errors. False–negative cytology has major medical, economic and legal implications, and this is reflected in high malpractice litigation costs in the USA associated with misreading cervical smears.

Testing for HPV with PCR is cumbersome and expensive. Current interest in the use of HPV DNA testing as a primary screening test is based on the finding that HPV DNA is present in almost all cervical cancers, and the availability of easily performed tests that have demonstrated higher sensitivity for high grade CIN (CIN2+) than that achieved by cytology. Primary screening with HC2 detects more than 90% of all CIN2, CIN3 or cancer cases, and is 25% (95% CI 15 to 36%) relatively more sensitive than cytology at a threshold of ASCUS (or low-grade squamous intraepithelial lesions [LSIL] if ASCUS unavailable), but is 6% (95% CI 4–7%) relatively less specific. The peak incidence of hrHPV infections in women occurs between 20 and 25 years of age, and most of these infections are cleared spontaneously. The sensitivity of HPV testing is independent of age, but specificity increases with age. The lower specificity of HPV testing in younger women is caused by a higher prevalence of transient infections. The higher positive predictive value (PPV) of both HPV testing and cytology in younger women, however, emphasises the much higher prevalence of CIN2+ lesions in this group.

The HART (HPV in Addition to Routine Testing) study found that the risk of women developing cervical cancer over the next 5–8 years was negligible if they tested negative on both hrHPV testing and cytology. In this cohort of 10,348 women, no woman with a positive initial HPV test or ASCUS smear developed CIN2+ lesions over the next 12 months. Human papillomavirus testing was more sensitive than Pap smear at ASCUS threshold (97.1% v 76.6%). These data formed the basis for the first recommendations that HPV testing could be the primary screening strategy, starting at age 30 years and repeating every 5 years if negative, and every 12 months if positive. The negative predictive value (NPV) has also been reported to vary between 97% and over 99% in other studies, and this high NPV has the potential to increase the screening interval to 5–8 years if this modality is used for primary screening. The American College of Obstetricians and Gynaecologists has recommended that co-testing using the combination of cytology plus HPV testing is an appropriate screening test for women older than 30 years. Any low-risk woman aged 30 years or older who receives negative test results on both cervical cytology screening and HPV DNA testing should be re-screened no sooner than 3 years subsequently. The proposed algorithm for primary screening using HPV testing as a starting point based on these various recommendations is shown in Fig. 1 .

Proposed algorithm using human papillomavirus deoxyribonucleic acid testing as primary screening and cytology as triage. DNA, deoxyribonucleic acid testing; HPV, human papillomavirus; NILM, negative for intraepithelial lesions or malignancy.

Naucler et al. used the database from the intervention arm ( n = 6257 women) of a population-based, randomised-controlled trial of double screening with cytology and HPV DNA testing to evaluate the efficacy of 11 possible cervical screening strategies based on HPV DNA testing alone, cytology alone, and HPV DNA testing combined with cytology among women aged 32–38 years. The main outcome measures were sensitivity for detection of CIN3+ disease within 6 months of enrollment or at colposcopy for women with a persistent type-specific HPV infection, and the number of screening tests and positive predictive value (PPV) for each screening strategy. All statistical tests were two-sided. The sensitivity, specificity, PPV and NPV respectively were 71.3%, 98.6%, 42.5%, 99.59% for cytology and 95.4%, 94.2%, 19.2% and 99.93% for HPV DNA testing. The number of screening tests required to detect one case of CIN2+ were 100.9 and 75.4 for cytology and HPV testing, respectively. Compared with screening by cytology alone, double testing with cytology and for type-specific HPV persistence resulted in a 35% (95% CI 15% to 60%) increase in sensitivity to detect CIN3+, without a statistically significant reduction in the PPV (relative PPV 0.76, 95% CI 0.52 to 1.10), but with more than twice as many screening tests needed. Several strategies that incorporated screening for high-risk HPV subtypes were explored, but they resulted in reduced PPV compared with cytology. Compared with cytology, primary screening with HPV DNA testing followed by cytological triage and repeat HPV DNA testing of women positive for HPV DNA with normal cytology increased the CIN3+ sensitivity by 30% (95% CI 9 to 54%), maintained a high PPV (relative PPV 0.87, 95% CI 0.60 to 1.26), and resulted in a mere 12% increase in the number of screening tests. They concluded that HPV DNA testing as primary screening followed by cytological triage and repeat HPV DNA testing of women who are HPV positive with intraepithelial lesions or malignancy after at least 1 year is the most feasible strategy in primary cervical screening.

The proportion of women aged over 30 years with normal cytology and testing positive for hrHPV varies by continent. In a meta-analysis of 1 million women with an overall prevalence of 11.7% (95% CI 11.6 to 11.7%), with the highest prevalence in developing regions of sub-Saharan Africa (24.0%), Eastern Europe (21.4%) and Latin America (16.1%). The proportion of women requiring triage by cytology or other methods and referral to colposcopy will accordingly vary. Although a large number of HPV positive women will be negative for intraepithelial lesions or malignancy (NILM), they are nevertheless at higher risk of developing an abnormal smear within the next 5 years. Castle et al. found that 15% of a cohort of 2020 women developed an abnormal Pap smear within 57 months. Rozendal et al. found that women who were HPV positive with NILM have a 116-fold increased risk of developing CIN3 over 40 months. By HPV testing, a subset of women is identified who will benefit from closer follow up, and the specificity of the test improves with time.

Restricting HPV screening to women aged over 30 years improves specificity by eliminating consideration of transient infections in younger women. It has been seen that testing women under 30 years increases the referral to colposcopy without increasing the detection of cases of invasive cancer under age 30 years. Other approaches under evaluation to deal with the lower specificity of HPV DNA testing include HPV genotyping of hrHPV-positive women with NILM or ASCUS cytology for HPV 16, 18 and 45, markers of proliferative lesions such as p16INK4a and mRNA coding for the viral E6, E7 proteins, or both, with a potential clinical use recommending closer follow up or earlier intervention in those who test positive. The specificity can also be improved by raising the threshold definition of a positive test. Cuzick et al. found that all positive cases among women aged 35 years and older had HPV DNA levels over 4 pg/mL and the sensitivity was 95.5%. The referral rate was 6.8% using the standard cut-off of 1 pg/mL but reduced to 4.2% at 2 pg/mL. In the New Technologies for Cervical Cancer trial, Ronco et al. conducted conventional cytology and HPV testing using the HC2 test in a multicentre randomised-controlled trial of 33,364 participants aged 35–60 years. They found that increasing the threshold from 1 pg/mL to 2 pg/mL did not substantially affect either the sensitivity or specificity (97.3% and 96.0%, 93.2% and 94.9%, respectively) but increased the PPV from 6.6% to 8.5%. They concluded that HPV testing alone was more sensitive than conventional cytology among women aged 35–60 years. Adding liquid-based cytology improved sensitivity only marginally but increased false–positives. Human papillomavirus testing using HC2 with a 2 pg/mL cut-off may be more appropriate than a 1 pg/mL cut-off for primary cervical cancer screening to improve the specificity and decrease referral to colposcopy. Other studies have also had similar conclusions. Gravitt et al. have cautioned that in scenarios where screening may be carried out only once or twice in a lifetime, it may be better to use the standard cut-off of 1 pg/mL to obtain the maximum possible sensitivity.

Using HPV DNA testing as the primary screening modality has several advantages: (1) HPV DNA detection assays provide an automated, objective, robust and sensitive test. This allows for better quality control and reduces the basis for medico-legal claims; (2) cytology can be reserved for the 5–15% of women who are HPV-positive. This facilitates high-quality cytology and allows the use of fewer, more focused cyto-screeners; (3) it also avoids the unnecessary triage of HPV-negative ASC-US and LSIL; and (4) a longer screening interval is likely to be safe, which would improve the cost and convenience of screening. In addition, adenocarcinoma and its precursors which, just like squamous-cell carcinomas, are nearly always hrHPV positive, are frequently missed by cytology, and hrHPV testing is helpful in detecting these lesions.

A major advantage of HPV testing is the potential for self-sampling, including home-based self-sampling. The method was first described from developing countries as a method for overcoming the problem of non-availability of resources for clinician sampling. Various types of self-samplers have been described, including the Digene ^® cervical sampler, Dacron swabs, conical brush and endocervical brush. Trials comparing self- and clinician sampling have been reviewed by Petignat et al. and further to this by Gravitt et al. The detection of high-risk HPV types has good concordance. Low-risk types are more often found in self-samples. For detection of CIN2+, the sensitivity of self-collected vaginal samples is consistently 10–19% less than that of clinician-collected cervical samples, but as sensitive as cytology. Difference in the specificity is only minimal. Apart from its potential role as a screening method in developing countries, self-sampling is being attempted as a method of home-based primary screening for underserved women in developed countries who do not access regular screening programmes and do not respond to invitations for screening. Screening participation has been found to increase four-fold by this method.