HPV and cervical cancer
Cervical cancer develops from hrHPV-infected cells that typically originate in the squamocolumnar junction, an area of high cell turnover. Through viral proteins E6 and E7, HPV causes reduced cell apoptosis and unregulated cell growth. The causal link between HPV infection and the development of cervical cancer was first identified in 1984 by Dr Harald zur Hausen, who won the Nobel Prize for isolating HPV types 16 and 18. Further study has identified 14 hrHPV types. HPV 16 and 18 remain the most prevalent types, and are associated with approximately 70% of all cervical cancers.
Not every woman who becomes infected with HPV develops a precancerous lesion or cancer. The majority of young women are able to clear the HPV virus. As evidenced by negative follow-up hrHPV DNA testing, approximately 90% of HPV infections are cleared within 1-2 years. Although it is possible that the HPV virus could remain dormant at undetectable levels, it is most likely that women make anti-HPV antibodies that confer long-term protection against subtype-specific infection. HPV infections that persist are at much greater risk of progression to a precancerous lesion or cervical cancer. When these lesions are detected, cervical excisional procedures are highly efficacious (90-95%) in eliminating preinvasive disease.
History of cervical cancer screening
In the 1920s, Dr George Papanicolaou began work on the cervicovaginal smear at Weill Medical College in New York City, which was first published with his partner Dr Herbert Traut in 1941. This test would later become known as the “Pap smear.” No large randomized control trials have confirmed its efficacy and as such, cervical cytology was never formally evaluated before being implemented as a screening test. However, global epidemiologic studies have convincingly demonstrated its efficacy as a cancer prevention strategy. In areas where cervical cytology was implemented, the mortality and morbidity from cervical cancer has been greatly reduced. In the United States, rates declined from 36.3 per 100,000 women in the 1930s to 7.2 per 100,000 women in the 1990s. The greatest decline was from 1950 through 1970, with a decline of 3% per year, which correlates with the adoption of routine cervical cancer screening programs in the United States ( Figure 1 ).
Traditional cytology has been shown to have a sensitivity of only 51% (30-87%) and specificity of 98% (86-100%). Additionally, due to the subjective nature of the screening test, there is significant interobserver variability in the interpretation of cytology, which further contributes to its variable sensitivity and specificity rates. Annual screening has been able to overcome this low sensitivity (ie, false-negative rate) of cervical cytology. A metaanalysis by Spence et al sought to understand how women develop cervical cancer despite availability of widespread screening programs. In this study, which examined 42 studies from 1950 through 2007, the authors concluded that an estimated 29% (95% confidence interval, 21–40%) of failures to prevent invasive cervical cancer can be attributed to false-negative cytology.
Given the inherent deficiencies of cytology screening, hrHPV testing developed initially as an adjunct to traditional screening. Approved by the Food and Drug Administration (FDA) in 1999, hrHPV testing was first recommended for reflex testing of atypical squamous cells of undetermined significance (ASCUS) cytology to triage patients to colposcopy. In 2004, the National Institute of Health National Cancer Institute, ASCCP, and ACS convened and agreed on interim guidelines to expand the use of hrHPV to cotesting women age ≥30 years given data that demonstrated high sensitivity and specificity in this age group. Many clinical trials have now been performed that have increased our understanding of the performance of hrHPV testing. The data have consistently shown HPV testing to have a higher sensitivity and reproducibility with increased negative predictive values compared to cytology, thus leading to a scientific and clinical interest in primary HPV testing as a method for cervical cancer screening.
History of cervical cancer screening
In the 1920s, Dr George Papanicolaou began work on the cervicovaginal smear at Weill Medical College in New York City, which was first published with his partner Dr Herbert Traut in 1941. This test would later become known as the “Pap smear.” No large randomized control trials have confirmed its efficacy and as such, cervical cytology was never formally evaluated before being implemented as a screening test. However, global epidemiologic studies have convincingly demonstrated its efficacy as a cancer prevention strategy. In areas where cervical cytology was implemented, the mortality and morbidity from cervical cancer has been greatly reduced. In the United States, rates declined from 36.3 per 100,000 women in the 1930s to 7.2 per 100,000 women in the 1990s. The greatest decline was from 1950 through 1970, with a decline of 3% per year, which correlates with the adoption of routine cervical cancer screening programs in the United States ( Figure 1 ).
Traditional cytology has been shown to have a sensitivity of only 51% (30-87%) and specificity of 98% (86-100%). Additionally, due to the subjective nature of the screening test, there is significant interobserver variability in the interpretation of cytology, which further contributes to its variable sensitivity and specificity rates. Annual screening has been able to overcome this low sensitivity (ie, false-negative rate) of cervical cytology. A metaanalysis by Spence et al sought to understand how women develop cervical cancer despite availability of widespread screening programs. In this study, which examined 42 studies from 1950 through 2007, the authors concluded that an estimated 29% (95% confidence interval, 21–40%) of failures to prevent invasive cervical cancer can be attributed to false-negative cytology.
Given the inherent deficiencies of cytology screening, hrHPV testing developed initially as an adjunct to traditional screening. Approved by the Food and Drug Administration (FDA) in 1999, hrHPV testing was first recommended for reflex testing of atypical squamous cells of undetermined significance (ASCUS) cytology to triage patients to colposcopy. In 2004, the National Institute of Health National Cancer Institute, ASCCP, and ACS convened and agreed on interim guidelines to expand the use of hrHPV to cotesting women age ≥30 years given data that demonstrated high sensitivity and specificity in this age group. Many clinical trials have now been performed that have increased our understanding of the performance of hrHPV testing. The data have consistently shown HPV testing to have a higher sensitivity and reproducibility with increased negative predictive values compared to cytology, thus leading to a scientific and clinical interest in primary HPV testing as a method for cervical cancer screening.
Delayed screening start until age 21 years
In 2009, experts convened and reviewed the available literature and recommended screening start at age 21 years. The rationale for this decision is based on the following primary concept: although the rates of HPV infection in the age group is high, the incidence of cervical cancer in this age group is exceedingly low. The Surveillance, Epidemiology, and End Results (SEER) data from 1998 through 2003 (and again in 2007 through 2011) estimated there was an average of 14 invasive cervical cancers diagnosed in US females <20 years old, with an incidence of 0.1 per 100,000 women. To prevent 1 case of cervical cancer, 1 million young women would need to be screened. This would lead to thousands of unnecessary colposcopies, biopsies, and excisional procedures. The costs and morbidities associated with this far outweigh the benefits in this population. To further emphasize the ineffectiveness of screening in this age group, an expert group compared SEER data from the 1970s to data from the present era, and noted that since the initiation of widespread screening programs in the United States, the incidence rate of invasive disease is largely unchanged in this age group indicating that screening in this population has had no effect on incidence.
Although the incidence of cervical cancer in young women is very low, many young women harbor both low-risk HPV and hrHPV infections. According to the 2003 through 2006 National Health and Nutrition Examination Survey, the US HPV prevalence in those <20 years of age is estimated at 32.9% with hrHPV prevalence estimated at 28.3%. Women age 20-24 years have slightly higher rates of hrHPV infection at 43.4%. From these new screening guidelines, one should not assume that women age <21 years do not acquire a persistent HPV infection or go on to develop a high-grade preinvasive lesion. Instead, because time to progression from preinvasive to invasive disease is long (years to decades), screening at age >21 years can still prevent the development of invasive disease.
Screening in the 20s and conservative management of low-grade cytology
Once screening has been initiated, one goal is to detect the majority of precancerous lesions while limiting morbidity associated with detection and treatment. The appropriate screening intervals and treatment guidelines in young women age 21-29 years slowly evolved throughout the early 2000s. Initially, the ACS, USPSTF, and American Congress of Obstetricians and Gynecologists (ACOG) wrote separate opinions throughout this time period with varying recommendations. In 2012, the ACS/ASCCP/ACS, USPSTF, and ACOG all aligned their recommendations to change the cytology screening interval to every 3 years. This increase in the screening interval was based on concepts similar to those that guided initiation of screening at age 21 years. Although this age group has a high rate of HPV infection (46.8-53.8%), rates of cervical cancer remain exceedingly low, at 1.2-1.4 per 100,000 in women age 21-24 years and 5.1 per 100,000 in women age 25-29 years. The appropriate screening interval should be set long enough to minimize harms, while still maximizing detection of abnormal lesions prior to progression. It should be noted that there is limited prospective data evaluating the appropriate screening interval in this young age group. A decision analysis by Kulasingam et al showed that the predicted death rate from cervical cancer with screening every 3 years was 0.05 per 1000 women. When this was compared to screening every 2 years, the predicted death rate was equivocal, but there was a 40% increase in the number of colposcopies performed.
As to whether to extend the interval >3 years, a large Scandinavian study reviewed the screening history of women with cervical intraepithelial neoplasia (CIN)3. This study found that women with ≥3 years since their last negative cervical cytology screen were at an increased risk of having CIN3 compared to women who had screening intervals <3 years with an age-adjusted incidence rate of 12.2 vs 1.5 and a relative risk of 1.3 (95% confidence interval, 0.6–3.2). Overall, the current 3-year guideline is based on limited data that are not specific to this age group, but to most providers, a 3-year interval appears to align with current acceptable risk-benefit parameters.
The most recent ASCCP guidelines recommend women age 21-24 years with ASCUS/HPV + or low-grade squamous intraepithelial lesion (LSIL) cytology results have repeat cytology at 12 months instead of colposcopy, which has been a past recommendation. The rationale for these changes is based primarily on the ASCUS-LSIL Triage Study indicating similar low risk of progression to cervical cancer with either LSIL or ASCUS/HPV + cytology in young women. Women age ≥30 years with LSIL cytology have a 5-year cumulative risk of CIN3 + of 5.2%, which is high enough to necessitate colposcopy examination. However, in women age <25 years, the risk is substantially lower at 3.0%. Although HPV prevalence is high in this age group, HPV clearance is also high and therefore delaying colposcopy until a woman has proven HPV persistence (with consecutive abnormal cytology) can be considered in younger women. However, given the possibility of persistence, these individuals should have closer follow-up to ensure clearance.
Extended screening interval in the 30s
According to SEER data from 2007 through 2011, compared to young women age <30 years, cervical cancer incidence is significantly higher in women ages 30-39 years at 24.0 per 100,000 women. The increased incidence necessitates a screening program that has a high sensitivity for early detection in this age group. It is also important to minimize testing of women at low risk of developing CIN3 + . Compared to younger women, women age ≥30 years are less likely to clear a new HPV infection and more likely to have HPV persistence. Therefore, the focus of screening is to detect persistent HPV infections and extend the screening interval testing for those who are HPV negative (and thus low risk).
In the most recent guidelines, women age ≥30 years can continue with cytology alone every 3 years. Notably, simply adding hrHPV cotesting to a 3-year screening interval would increase the detection rate of HPV, but would simultaneously lead to an increase in colposcopy examinations. As demonstrated in the model by Kulasingam et al, this would translate to an increase from 3-15 incremental colposcopies per life-year. However, studies have shown the safety and cost-effectiveness of extending the screening interval in this age group to every 5 years with a negative HPV cotest. A European study of 330,000 women demonstrated that women with negative cytology had a risk of CIN3 + of 0.17% at 3 years. When women were HPV negative, either via primary HPV test or with a cotest, the rate of CIN3 + at 5 years was similar at 0.17% and 0.16%, respectively, indicating comparable risk to screening with cytology every 3 years. Katki et al used the large database of HPV cotests collected by Kaiser Permanente Northern California to demonstrate that the 5-year cumulative incidence rate of CIN3 + in women age ≥30 years with negative cytology screening is 7.1 per 100,000, but with a concurrent negative HPV cotest that rate was reduced by >50% to 3.2 per 100,000. Thus, given acceptable reductions in risk with minimal impact on the harms, the ACS/ASCCP/ASCP Consensus guidelines recommended cytology with HPV cotesting for women age ≥30 years at 5-year intervals. As Kinney et al recently argued, the safety of extension to a 5-year interval was based on comparison to 3-year screening intervals and not to annual screening cytology. This indicates a potentially higher risk of a subsequent cervical cancer diagnosis long-term compared to the gold standard of annual cytology. Moving forward, the risk of cervical cancer and potential death (not just CIN3) should be considered in future modeling studies.
Another important benefit of HPV cotesting is the diagnosis of cervical adenocarcinoma and adenocarcinoma in situ. Due to its endocervical location, traditional cytology often fails to diagnose these lesions. In a large pooled analysis of almost 200,000 women enrolling in various HPV screening trials, HPV testing (compared to traditional cytology) showed improved detection of adenocarcinoma compared to squamous cell carcinoma. Adenocarcinoma incidence is increasing and accounts for 15-25% of invasive cervical cancers, further supporting the addition of HPV testing in screening algorithms.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
