A 21-Year-Old Woman With Atypical Squamous Cells of Undetermined Significance

DR REYNOLDS: Ms G is a 21-year-old woman with a history of abnormal Papanicolaou (Pap) tests.

Ms G first became sexually active at age 19. That year, she had her first gynecologic examination when she visited a doctor to ask for an oral contraceptive pill prescription. Her examination and cytology test were normal. She had annual Pap screening until age 21, when her cytology test (conventional) showed atypical squamous cells of undetermined significance (ASC-US). Her clinician, a nurse practitioner at student health, recommended a repeat cytology test in 6 months. Her repeat test again showed ASC-US, and she was referred to a gynecologist who performed a colposcopy.

On colposcopy, the entire squamocolumnar junction could be seen and her cervix showed grade 1 acetowhite changes. Biopsy revealed metaplasia. A liquid-based cytology test done at the time of the colposcopy was normal. Testing for human papillomavirus (HPV) DNA was not performed.

Ms G has never been diagnosed with a sexually transmissible infection. She has never been pregnant. She does not take medications or smoke tobacco. She is a recent college graduate and is working as a writer.

I was only sexually active about 4 months when I had my first Pap test. I was 19. It was my sophomore year in college—the Pap test was normal and everything was fine. Then, when I was 21, I had the first abnormal Pap, and then a repeat abnormal Pap. They both came back as ASC-US.

At first, when I got the results, I was totally freaked out because I just figured that I was going to skate by as usual. They don’t tell you when you get a Pap what could cause an abnormal test. After I got the results, I was on the Internet reading stuff, and I was reading books and trying to figure out what had happened. It was a little disturbing.

It did affect the way that I feel about both pelvic exams and sexual activity because I was forced to look back at myself and see what aspects of my behavior could have brought this on. But then, once it became more familiar to me and I could talk about it with my friends and my peers, I realized how common abnormal Paps and colposcopies are. We just never talked about it because no one brought it up. So, as far as it making me wary of sexual partners in the future, I think I’ve always been careful. I don’t think this has really changed anything. It’s just made it more realistic as to why I have to be so careful.

My nurse practitioner and physician were very thorough and told me that ASC-US is common, what causes it, and how it can just simply be from a bad reading of the slide. The doctor said that I should repeat the Pap test every 3 to 6 months. That’s a comfortable window for me.

This experience has made me much more likely to get Pap tests. Not only have I encouraged my friends to get Paps, I’ve badgered them. The ones that were afraid to go to the doctor have now seen 3 of our good friends have very similar experiences and have seen why it’s important to catch it in the very beginning so that you can take care of it properly.

I would like to know why clinicians don’t discuss what the abnormal results might be before a Pap test is performed. I think that would make women less flippant about how important a Pap smear is.

When should cervical cancer screening begin? Are liquid-based cytology tests better than conventional tests? What is ASC-US, and what does it indicate? Is an HPV DNA test an appropriate next step in the triage of an ASC-US result for all women? Once a patient has been evaluated for ASC-US, what should the next steps be? What do you recommend for Ms G?

DR SAWAYA: Within her first 3 years of being screened, Ms G has had 6 cervical cytology tests, a colposcopy with a cervical biopsy, and 2 more cytology tests are planned within the coming year. So far, no cervical disease has been identified. What can we learn from her case, and how might we advise her?

Extensive cervical cancer screening in the United States has been accompanied by a striking decline in squamous cell cervical cancer incidence over the last several decades.¹ Incidence varies with age, ranging from 1.3 per 100 000 in women aged 20 to 24 years to 14.8 per 100 000 in women older than 50 years.^{2 - 3} Cytology-based screening involves cervical sampling to determine the presence of abnormal dysplastic cells. If abnormalities are present, a magnified view of the cervix (colposcopy) is often performed to identify and biopsy preinvasive, cancer precursor lesions known as cervical intraepithelial neoplasia (CIN). Cervical intraepithelial neoplasia is graded by the proportional thickness of the epithelium involved with abnormal cells, ranging from grades 1 to 3. Cervical cancer prevention is achieved by removing CIN grades 2 and 3 (high-grade dysplasia) with either excision (eg, cone biopsy, loop excision) or destruction (eg, cryotherapy, laser). The most common squamous cytological abnormalities in order of their strength of association with underlying CIN grade 2+ are: (1) ASC-US, (2) low-grade squamous intraepithelial lesion (LSIL), and (3) high-grade SIL (HSIL). While cytology-based screening programs have been successful in reducing cervical cancer, cytology has been criticized for poor reproducibility⁴ and inaccuracy, especially of ASC-US interpretations.

Atypical squamous cells of undetermined significance is defined as squamous cell abnormalities (eg, nuclear enlargement, slight hyperchromasia) not fulfilling strict criteria for dysplasia.⁵ Atypical squamous cells of undetermined significance is the most common cytological abnormality in women of all ages and is interpreted in 3% to 8% of all tests.⁶ More than 2 million women are diagnosed annually in the United States,⁷ approximately 10% of whom can be expected to have underlying CIN grade 2+.^{7 - 8} In the 2001 Bethesda System, the ASCUS category was redefined as “atypical squamous cells” and subdivided into “atypical squamous cells—cannot exclude high-grade SIL” (used to describe atypical cells at higher risk of being associated with CIN grade 2+) and “atypical squamous cells of undetermined significance” (ASC-US).⁵ In the following discussion, the term “ASCUS” will be used when referring to studies that used the older definition and “ASC-US” will be used to describe contemporary cytological interpretations.

The decision to begin cervical cancer screening is of greater importance than clinicians often appreciate and, in Ms G, began earlier than is currently recommended. Evidence from observational studies suggests that screening should not begin before the onset of vaginal intercourse, but more important, not too soon thereafter. The American Cancer Society (ACS)⁹ and the American College of Obstetricians and Gynecologists (ACOG)¹⁰ suggest that screening be delayed for approximately 3 years, but begin no later than age 21 years; recommendations by the US Preventive Services Task Force are similar¹¹ (Table). This upper age limit is designated to ensure that women unable or unwilling to disclose previous sexual intercourse are protected. The recommendation to not begin earlier is based largely on the following lines of evidence: most cervical cancer is caused by infection with high-risk HPV types¹² ; cervical-vaginal HPV prevalence is less than 2% before initiation of vaginal intercourse,^{13 - 15} and cumulative incidence within 24 months of coitarche is about 40%.¹⁵ Acute HPV infection causes cervical cell changes that can manifest as low-grade abnormal cytology, but most such cytology does not indicate the presence of underlying CIN grade 2+^{7 ,16} ; many cytological abnormalities¹⁷ and some CIN grade 1 and 2 lesions^{16 ,18} will resolve spontaneously and never be clinically relevant (except in the sense that they may lead to further interventions if detected during screening). Finally, cervical excisional treatments, which may eventually result from positive screening, have been associated with adverse obstetrical outcomes.^{19 - 20}

The ACS summarizes its recommendations as follows: “Screening before the 3-year period may result in an over diagnosis of cervical lesions that will regress spontaneously leading to inappropriate interventions that may do more harm than good.”⁹ Ms G reported having her first cytology test within only a few months of becoming sexually active and 2 years before it was recommended. As it turns out, this was an imprudent clinical decision that set in motion a series of interventions and investigations that have yet to yield a beneficial health outcome.

Assuming a patient is of appropriate age and wishes to be screened, there are 2 cytology-based screening options. Liquid-based cytology (LBC) was developed largely to improve the quality and adequacy of cervical specimens compared with that of conventional cytology with an ultimate goal of facilitating slide interpretation.²¹ Liquid-based cytology samples are collected in a manner similar to conventional cytology, but instead of being placed onto a slide and fixed, the specimen is placed in a preservative liquid medium. In the laboratory, processing removes debris and places a thin layer of cells on slides that are then stained and interpreted similarly to conventional tests.

Although the data comparing LBC and conventional cytology are voluminous, the US Preventive Services Task Force concluded: “Overall, the quality of this literature is poor for the purposes of making decisions about choice of screening systems in US populations.”¹¹ Many studies have been case series with either nonrandomized controls or with historical controls, study designs with inherent biases.²² In the only published randomized comparison,²³ women screened with LBC did not differ in cytological abnormalities or the number of CIN grade 2+ cases identified compared with those screened with conventional cytology. Given the sample size (N = 1999), however, the study had limited power to determine these differences. The proportion of wholly unsatisfactory tests was higher in the LBC group (1.4% vs 0%). The cytology collection method and interpretations used in this study, however, may not be generalizable to other settings.

The distinction between LBC and conventional cytology is important because the ACS guidelines recommend that screening be performed annually with conventional cytology or every 2 years with LBC (Table).⁹ This differential recommendation is not based on potentially greater LBC sensitivity, but on potentially poorer LBC specificity. The guideline cautions that annual screening with LBC will “likely lead to increases in the detection of ASC-US and low-grade abnormalities with subsequent increases in referral to colposcopy unnecessarily, risking the potential for overtreatment and increased health care costs.”⁹ Of relevance to Ms G, ASCUS rates of more than 10% and overall cytology positivity rates of 20% have been reported with LBC in women of her age.²⁴

In 2002, the American Society for Colposcopy and Cervical Pathology (ASCCP) published guidelines²⁵ stating that colposcopy, repeat cytology, and DNA testing for high-risk HPV types are all acceptable ways to manage a patient with an ASC-US test result. An ACOG practice bulletin published in 2005 makes similar recommendations.²⁶ The best evidence about which approach to choose comes from the ASCUS/Low-grade SIL Triage Study (ALTS), a large, multicenter randomized trial sponsored by the National Cancer Institute.²⁷ In this study, 3488 women (mean age, 29 years) with a single community-interpreted ASCUS test were randomized to receive immediate colposcopy; repeat cytology every 6 months with subsequent colposcopy in women with cytology interpreted as HSIL or worse (HSIL+); or testing for high-risk HPV DNA with subsequent colposcopy in women testing positive. The HPV DNA testing was performed with a high-risk probe set (Hybrid Capture 2 [HC2], Digene Corporation, Gaithersburg, Md [HC2 detects 1 or more of the 13 high-risk HPV types: 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68]).

Based on the results of the group randomized to receive colposcopy, 11% of women with a single ASCUS test had underlying CIN grade 2+ (CIN grade 2, 6.3%; CIN grade 3, 5.1%).⁷ At enrollment, about 56% of women with ASCUS had a positive high-risk type HPV DNA test, and nearly all prevalent CIN grade 3+ cases (96%) were identified among the HPV DNA–positive group. Therefore, HPV DNA testing has a high sensitivity for CIN grade 2+, with about half of women so tested being referred to undergo colposcopy. Since colposcopy can be painful and is expensive, as an initial step in evaluation of a patient with an ASC-US test, HPV DNA testing is clearly an attractive alternative to colposcopy.

However, since less than 20% of obstetrician-gynecologists perform colposcopy for a single ASC-US test result,²⁸ the more clinically relevant question is the comparison between HPV DNA testing and repeat cytology.²⁹ During the 2-year study, strategies sending greater proportions of women to undergo colposcopy early (immediate colposcopy and HPV DNA testing) identified more cases of CIN grade 3+ sooner than did repeat cytology followed by colposcopy for women who had an HSIL+ result.¹⁶ Diagnostic delay in the repeat cytology group was not associated with more cervical cancer but did identify fewer cases of CIN grade 2, suggesting that these lesions regress over time. At study end, repeat cytology with colposcopy for an HSIL+ result referred about 12% of women to colposcopy compared with about 56% of women in the HPV DNA testing group.

While these results are important, current clinical guidelines²⁵ recommend, and more than 95% of clinicians²⁸ use, ASC-US+ rather than HSIL+ as the threshold for colposcopy. The ALTS did not directly evaluate ASC-US+ as the referral threshold, but had such a threshold been used, given that 58% of women had an ASCUS+ cytology test at study enrollment,¹⁶ more women would have been referred to undergo colposcopy. Assuming a similar likelihood of ASCUS+ if repeat testing had been done at 6 months, colposcopy referral rates at an ASCUS+ threshold appear similar to rates performed for positive HPV DNA testing; more cases of CIN grade 3+ would likely have been found sooner than with the HSIL+ referral threshold used in the ALTS.

In a systematic review of cross-sectional studies comparing HPV DNA testing for high-risk types to repeat cytology in managing ASCUS, HPV DNA testing had greater overall sensitivity (94.8%) and similar specificity (67.3%) compared with repeat cytology performed at an ASCUS+ threshold (sensitivity, 81.8%; specificity, 57.6%).⁸ However, cross-sectional studies should be cautiously extrapolated to clinical settings in which women are managed prospectively, since the strategies in these studies used CIN grade 2+ as the threshold for disease and CIN grade 2 is known to regress over time. Therefore, in a cross-sectional study, strategies in which women undergo colposcopy sooner (immediate colposcopy and HPV DNA testing) will identify more CIN 2 and appear more sensitive than a “waiting” strategy of repeating cytology in 6 months.

Is an HPV DNA test an appropriate next step in the triage of an ASC-US test result for all women, including those of Ms G’s age? Positivity rates for HPV DNA with ASCUS vary greatly with age, ranging from 71% in women aged 18 to 22 years to 31% in women older than 29 years in the ALTS.³⁰ In women older than 40 years, 20% with ASCUS were HPV DNA positive. Therefore, average ALTS results should be applied cautiously to women in these extreme age groups. In other settings, HPV DNA positivity rates with ASCUS in women younger than 30 years range from 49%³¹ to 80%.³² According to the ALTS investigators, a triage test that results in colposcopy for 80% or more of women would be impractical; the LSIL portion of the trial randomizing women to receive HPV DNA testing was stopped early when more than 80% of women were noted to be HPV DNA positive.³³

In summary, although both HPV DNA testing and repeat cytology at an ASC-US threshold are sensitive, less invasive alternatives to colposcopy, both strategies have suboptimal specificity and substantial proportions of women undergo colposcopy who do not have CIN grade 2+. How should a clinician decide between these 2 strategies?

As with all medical decisions, each strategy has trade-offs. If HPV DNA testing can be done without the patient having to return (ie, from the LBC vial or from a separate stored cervical swab collected at the time of conventional cytology), only those with an ASC-US test result and a positive HPV DNA test (about half) will need to return for colposcopy. About 60% of these women, however, will have no evidence of cervical neoplasia,³³ and the optimal surveillance strategy for them is unclear.²⁵ Follow-up analyses from the ALTS suggest that HPV DNA testing at 12 months may be an optimal management strategy for these women,³⁴ but an estimated 55% will remain HPV DNA positive³⁴ and, as per ASCCP²⁵ and ACOG²⁶ guidelines, would require another colposcopy. Since the generally accepted treatment threshold is CIN grade 2+,³⁵ some women with CIN grade 2 that is destined to resolve without clinical consequence would be unnecessarily treated and exposed to treatment harms. In addition, women given information that HPV is a carcinogenic virus with no known treatment or cure may have adverse psychological sequelae.^{36 - 37} They may also be concerned about recommendations stating they should disclose their HPV positivity to partners before having sex.³⁸

The repeat cytology strategy requires that all women return for cytology testing in 6 months. About half can be expected to have an ASC-US+ result and must return yet again for colposcopy. Return visits are inconvenient and costly. As with HPV DNA testing, most of these women will not have cervical neoplasia diagnosed at colposcopy. Waiting 6 months has the benefit of allowing CIN grade 2 to resolve with no evidence of cancer risk increasing over this short surveillance period, but it also enhances the likelihood that women will not return at all.

In medical situations such as this in which the evidence does not support one management strategy as clearly superior and the trade-offs involve patient-centered issues such as convenience and preferences for HPV DNA testing, clinicians should engage women in shared, informed decision making. Women in both research³⁹ and clinical⁴⁰ settings express a wide range of preferences for the 3 ASC-US management options. Integrating an individual woman’s preferences⁴¹ and values into clinical decision making is especially important when testing for medical conditions that may have adverse psychosocial ramifications.⁴² While ordering “reflex” HPV DNA testing from the LBC vial may be a convenient option for clinicians, testing without a woman’s permission and knowledge is inappropriate and violates ethical principles by not respecting patient autonomy in decision making.⁴² While women should be aware that cervical cancer screening aims to identify and destroy HPV-related lesions that have a high likelihood of progressing to cancer, clinicians should not be cavalier about testing for HPV DNA and providing positive HPV DNA test results to women who do not desire such information.

From the population perspective, cost-effectiveness analyses suggest that HPV DNA triage of ASC-US has cost advantages over both colposcopy and repeat cytology.^{43 - 44} These analyses, however, are limited by lack of data on longer-term screening outcomes, particularly in women with positive HPV DNA testing but normal colposcopy. Decision analytic studies using outcomes from the ALTS should help to delineate which ASC-US management strategy optimizes trade-offs in terms of sensitivity (cases of CIN grade 2+ found and estimated cervical cancer cases averted), specificity (women referred for positive tests who do not have a CIN grade 2+ result), and the inconvenience and expense of returning for additional testing. Cost and quality of life considerations will be important components of these analyses. The optimal management strategy may well vary depending on age. Testing for HPV DNA, for example, may be the “best” option in women older than 30 years; HPV positivity rates are substantially lower in this age group and sensitivity remains high.³⁰ Of importance, a recent pan-Canadian panel recommended HPV DNA triage for ASC-US, but only in women older than 30 years.⁴⁵

Both the ASCCP²⁵ and ACOG²⁶ guidelines suggest that if HPV DNA testing is used, those who test negative should have repeat cytology in 12 months; women who test positive should have colposcopy (Figure). The ACOG practice bulletin further states that as an alternative to colposcopy, adolescents with ASC-US and a positive HPV DNA test may be monitored with cytology at 6 and 12 months or with a single HPV test at 12 months.²⁶

Ms G began having cervical cytology tests within a few months of becoming sexually active, coincident with her request for oral contraception pills. Given her high likelihood of HPV infection and subsequent low-grade cytological abnormalities of no ultimate clinical importance, screening her at this time is a poor choice. Unfortunately, initiation of screening is commonly performed at this time.⁴⁶ Although Ms G would have been screened for the first time at age 21 years under ACS guidelines and presumably would have experienced the same cascade of testing if that was her first screening, her case illustrates the downsides of screening young women within 3 years of onset of sexual activity. In this age group, well-woman care should focus on family planning, screening for treatable sexually transmissible infections, and counseling about safe behaviors. Cervical cytology is commonly required before provision or refilling of hormonal contraception⁴⁷ but should certainly not be a prerequisite.⁴⁸ If her first 3 cytology tests had been liquid-based, then ACS recommendations suggest biennial, not annual screening; frequent screening has a greater likelihood of identifying transient changes in cervical cells of no clinical importance, as Ms G’s case illustrates.

All 3 of the options for ASC-US management have severe limitations in women of Ms G’s age. If faced with Ms G’s situation with an ASC-US test, I would present to her all options, but I would not recommend immediate colposcopy and, like her clinician, I would recommend repeat cytology. If her initial cytology had been liquid-based, HPV DNA testing would be simple to perform on the existing sample. Although this is the strategy preferred by ASCCP, it seems suboptimal in this age group since 70% to 80% of women with an ASC-US test result will be HPV DNA positive and would be referred for colposcopy. If her ASC-US test had been with conventional cytology and co-collection for HPV DNA testing had not been done, I would not have her return solely for the HPV DNA test since one of the great benefits of HPV DNA testing is not requiring the patient to return.

Repeat cytology in 6 months with an ASC-US threshold of referral has good sensitivity for CIN grade 3+, minimizes the likelihood of diagnosing transient CIN grade 2, and avoids both the potential adverse psychosocial aspects of HPV DNA testing and the establishment of the clinically troublesome “gray zone” of women who are HPV DNA positive but have no colposcopic evidence of cervical neoplasia. However, repeat cytology is no panacea; up to two thirds of young women will have ASCUS+ cytology 6 months later³⁰ and will need colposcopy as per ASCCP²⁵ and ACOG²⁶ guidelines. An ill-defined proportion with normal colposcopy will have persistent abnormal cytology at 12 months. If Ms G desired either immediate colposcopy or HPV DNA testing, I would counsel her about benefits and harms and ultimately respect her informed choice given that all 3 options are deemed appropriate by both ASCCP²⁵ and ACOG²⁶ guidelines. If HPV DNA testing is preferred by the patient and testing is positive, ACOG’s alternative guidelines for adolescents may be considered; instead of immediate colposcopy, HPV DNA testing can be repeated in 1 year, reserving colposcopy for those testing persistently HPV DNA positive. The risk of missing occult invasive cancer is low in young women, and those resolving HPV infections over a year’s time will be spared colposcopy.

Triage strategies that have improved specificity are clearly needed, especially for young women like Ms G with ASC-US. Recent evidence from the ALTS suggests that HPV DNA triage specificity may be greatly enhanced by type-specific HPV testing. In the ALTS, 63% of women younger than 30 years with ASCUS tested positive for 1 or more of 13 high-risk HPV types. HPV type 16, however, was positive in only 19% and predicted a greater absolute 2-year risk of CIN grade 3+ (32% vs 16%).⁴⁹ Future triage tests that narrow the group of women who are HPV DNA positive but who have no colposcopic evidence of CIN would be of great clinical value,⁵⁰ especially if they simultaneously define a high-risk subgroup. Food and Drug Administration–approved type-specific HPV tests are not yet available for use in the United States but may prove to be particularly useful as triage tests in young women.

Ms G expressed the desire to know more about screening goals and outcomes at the time of the initial examination. Educating women about the viral etiology of cervical cancer, the potential role of HPV DNA testing, screening benefits and harms and common testing outcomes, particularly for this age group, may best be done at this time. The ACS has informative educational materials available on its Web site (http://www.cancer.org/docroot/home/index.asp).

Ms G had a normal colposcopy and is having cytology tests every 6 months for 18 months. Given that no cervical neoplasia was found, both ASCCP²⁵ and ACOG²⁶ guidelines suggest she wait a year before repeat cytology (Figure). The majority of women in this age group managed with repeat cytology will likely fall in this category. Repeat cytology seems to be a safe and expeditious way to avoid intensive surveillance in young women with satisfactory normal colposcopy. Ms G does not smoke, but in our clinic, we advise all women to avoid cigarette smoking and we provide access to formal cessation programs because of the known association between smoking and increased cervical neoplasia risk⁵¹ and for the promotion of healthy behaviors.

While Ms G’s experience with cervical cancer screening is unfortunate, it has paradoxically strengthened her belief in the importance of screening such that she has actively promoted testing among her colleagues. Young women enthusiastic about cervical cancer screening need to be made aware of the projected benefits and potential harms of screening and treatment. In addition, professional organizations involved in setting screening policy for young women need to make clear, explicit statements that actively discourage screening too soon.

Ms G’s case illustrates that cervical cancer screening in young women can, and often does, elicit anxiety and a cascade of clinical interventions of no clinical value. Until we have more specific screening and triage tests, our best defense against the diagnostic dilemma posed by ASC-US in young women is “primary prevention of uncertainty” by exercising restraint and prudence in screening initiation.

QUESTION: These very well done studies such as the ALTS are using a surrogate end point rather than mortality from cervical cancer. I think if you had another ALTS with a fourth arm where no patient got screened until age 30 and the end point of the trial was cervical cancer mortality, I would bet that there would be no significant difference between any of those 4 arms.

DR SAWAYA: Cervical cancer mortality would certainly be more clinically important but is too rare to study prospectively. We are left with relying on surrogate markers that we think are the clearest indicators of increased risk and are unlikely to regress, such as CIN 3. We then must get comfortable with mathematical modeling studies that project forward the likelihood of CIN 3 progressing to invasion over discrete periods of time and how changes in screening strategies affect cancer incidence and mortality. The role of modelers is vital, and it is imperative that they be objective, clinically savvy, and in tune with the latest evidence. Once you start putting modeling in the equation, however, people get either confused or suspicious, and they have reason to be a little of both. I think the ALTS was an important trial that validated HPV DNA triage as a viable strategy for managing ASC-US compared to colposcopy in everyone, but it is important to note that the ALTS was designed primarily to determine which triage strategy strikes the optimal balance between finding maximal cases of high-grade CIN while at the same time minimizing resource expenditures and screening harms. That said, cervical cancer incidence might differ between strategies if substantial proportions of women managed in certain ways do not follow up for diagnosis and treatment.

QUESTION: One of the problems that has occurred for the clinician is the lumping together of CIN 2 and CIN 3 as “high grade.” High-grade dysplasia must be treated, but as you’ve pointed out, some CIN 2 will go away even if we don’t treat it. We would be much better off if we stayed with CIN 2 and CIN 3 so we can better assess which patients should be treated.

DR SAWAYA: I agree. This distinction is especially important since some experts at the ASCCP consensus conference expressed the opinion that observation of CIN 2 rather than treatment is appropriate in properly counseled, reliable adolescents.³⁵ The ACOG guidelines also endorse observation as a reasonable management strategy in adolescents with CIN 2.²⁶

QUESTION: Is your algorithm the same in a pregnant patient with ASC-US?

DR SAWAYA: The ASCCP guidelines state that we should manage cytological abnormalities in pregnant women the same as nonpregnant women.²⁵ The ACOG guidelines state that colposcopy may be deferred 6 to 12 weeks postpartum in pregnant women of all ages with cytology results of ASC and LSIL.²⁶ But we should rethink the value of cervical cancer screening at all in pregnant adolescents. Look at the evidence: we know that all pregnant adolescents have had unprotected intercourse, so the likelihood of HPV exposure is high, and cytology abnormalities such as ASC-US and LSIL are common. If we do cervical cytology as a part of routine prenatal care and follow ASCCP guidelines, we will do colposcopy in a substantial proportion of these women during their pregnancies. Remember, however, that we do screening and colposcopy in pregnancy to identify occult invasive cancer, not CIN. We don’t treat CIN in pregnancy, but a diagnosis of invasive cancer has implications for a patient’s pregnancy and delivery plan. Consider the trade-offs: you have a cytology test that’s positive in up to 1 in 6 women under the age of 20 years,²⁴ but the prevalence of cervical cancer in this age group is about 1 per million.³ It doesn’t take complex mathematical modeling to conclude that this is an unwise health policy. I personally do a lot of colposcopy in these women and find this trade-off unsatisfying. Although we have no direct evidence, I am concerned that the experience of colposcopy in pregnancy may discourage some young women from attending clinic for other important health services in the future. For this reason, our clinic guidelines state that cervical cancer screening should be discouraged in pregnant women under age 20 years.

QUESTION: I am a general internist and a male patient in his 50s asked me recently, “Should I have HPV testing? I’m afraid I’m infecting my sexual partners.”

DR SAWAYA: The clinical utility of HPV testing in men, especially for screening, is obscure. First of all, there is no Food and Drug Administration–approved test for HPV for men, and second, the counseling and optimal clinical management of the average man found to be HPV positive is unclear. Although condom use has been shown to hasten HPV regression in women,⁵² high-quality evidence about the effectiveness of condoms in primary HPV transmission is lacking.⁵³ Moreover, the time horizon of infectivity is unclear. While knowledge of carcinogenic HPV DNA positivity may well guide screening decisions in women over age 30 years, the value of this knowledge in other settings is unclear. We live in a society of sexually active people, at least 50% of whom will acquire genital HPV infection in their lifetime.⁵⁴ For this reason, I downplay the significance of HPV positivity except in the context of cervical cancer screening. In general, we need to be thoughtful about the clinical utility and psychosocial implications of many newer tests. Just because we can test doesn’t mean that we should test.

QUESTION: Do you think HPV vaccination will ever make cervical cancer screening obsolete?

DR SAWAYA:Prophylactic vaccines targeting carcinogenic HPV types 16 and 18, the types most frequently implicated in cervical cancer, have been evaluated in randomized trials^{55 - 57} and results are encouraging. At least 16 other HPV types, however, have been implicated in cervical carcinogenesis.¹² If we do vaccinate successfully against type 16 and 18 and the prevalence of dysplasia secondary to these 2 types falls, then the specificity of our screening tests has to go up remarkably or else we will have extremely poor positive predictive value. Though the prospect of eventually decreasing cervical cancer incidence through primary prevention of HPV infections is real, we have a lot to learn about vaccination and its ultimate impact on dysplasia and cancer. Remember that most sexually active women have already been exposed to HPV and efficacious therapeutic, rather than preventive, vaccines have yet to be validated. For this and other reasons, I believe that in our lifetimes we will always have some sort of cervical cancer screening in the United States.

QUESTION: It appears that the harms of cervical cancer screening for most women under 25, and many women under 30, may outweigh the benefits. Is there any movement toward sharing that information with patients and allowing them a choice in cervical cancer screening?

DR SAWAYA: The recent ACS and ACOG recommendations endorsing delay of screening onset are an incremental step in the direction of trying to maximize benefits and minimize harms by doing less screening in low-risk women. I think it would be reasonable for an informed young woman who does not want cervical cancer screening to simply opt out. I am a strong advocate of integrating patients’ preferences and values into clinical decision making, so I hope that expression of such preferences will not pose barriers to other important preventive care, such as provision of contraception.