brand logo

Am Fam Physician. 2006;73(4):719-729

Cervical cytology screening has been associated with a dramatic reduction in cervical cancer incidence and mortality. Current screening techniques may result in unnecessary visits, procedures, and patient anxiety; however, the value of accurate screening results can be reduced by loss to follow-up or undertreatment of significant lesions that may progress to invasive cancer. To address these issues, the American College of Obstetricians and Gynecologists (ACOG) has released evidence-based guidelines for management of abnormal cervical cytology and histology. The recommendations were published in the September 2005 issue of Obstetrics & Gynecology.

The Bethesda System should be used to communicate accurately the risk of cervical intraepithelial neoplasia (CIN) grades 2 and 3, adenocarcinoma in situ (AIS), or cancer (collectively, CIN 2-3+). If the cytology results do not define that risk clearly because of the use of categories not found in the Bethesda System, the physician may wish to request an interpretation that falls within the Bethesda System from the interpreting laboratory or cytopathologist.

The natural history of CIN is linked to the presence of high-risk human papillomavirus (HPV). Carriage of HPV DNA is extremely common in the general population; infection occurs at a reported rate of 1.2 to 1.3 percent per month. Extrapolating from these rates, the lifetime cumulative risk is at least 80 percent. Most women clear the virus or suppress it to levels not associated with CIN 2-3+. The duration of HPV positivity is shorter and the likelihood of clearance is higher in younger women.

Only one in 10 to one in 30 HPV infections is associated with abnormal cervical cytology results, with an even smaller proportion associated with CIN 2-3+. Among women with negative cytology results and a positive HPV test result, only 15 percent will have abnormal cytology results within five years. However, HPV is necessary for the development and maintenance of CIN 3. Persistent high-risk HPV infection is necessary for the development of almost all invasive cancers. Conversely, the risk of cervical cancer in women who are not infected with high-risk HPV is extremely low.

From a clinical perspective, it is important to determine which intraepithelial neoplasias will progress to invasive cancer if left untreated. However, the diagnostic categories currently available have only modest predictive value, and that value decreases as the lesions become less severe. Cancer precursors include CIN 3; AIS; and, to a lesser extent, CIN 2. The likelihood of progression to cancer is higher and the time to progression is shorter as the grade of dysplasia increases.

Cervical Cytology

Cervical cytology screening is associated with a reduction in the incidence of and mortality from invasive squamous cancer. Conventional cytology is reported to be 30 to 87 percent sensitive for dysplasia; a meta-analysis of cervical cytology studies suggested a sensitivity of 58 percent in one screening population. Treatment based on conventional cytology results does not seem to decrease the incidence of glandular invasive cancers, suggesting that sensitivity for glandular precursors is less than that observed for squamous lesions. Because the range of sensitivity (30 to 87 percent) is so broad, all abnormal cytology results must be evaluated, although most do not represent underlying CIN 2-3+. When cytologic testing is selected for follow-up of previous abnormal results, repeat testing at six- to 12-month intervals is recommended.

HPV Testing

Testing for low-risk HPV types has no role in cervical cancer prevention. The low-risk HPV types are associated with genital warts and low-grade intraepithelial lesions of the cervix, vagina, and vulva.

For women 30 years and older, HPV testing can help predict whether CIN 2-3+ will be diagnosed in the next few years in those who have a normal cytology result. Results are similar between hybridization and polymerase chain reaction methods if the positive cutoff and viral types tested for are similar. As new tests are introduced, decisions about clinical practice implementation must be based on clinical sensitivity (i.e., relationship of the test result to CIN 2-3+), not analytic sensitivity (i.e., ability of the test to detect low levels of HPV).



Women with a normal cervical cytology result who test positive for HPV on routine screening have an approximately 4 percent risk of developing CIN 2-3+, which is lower than the risk for women with atypical squamous cells (ASC). For this reason, colposcopy is not recommended as further testing after a single HPV-positive, cytology-negative result. Current expert consensus guidelines recommend repeat cytologic and HPV testing in six to 12 months to allow for resolution of transient HPV infection and colposcopy only if test results remain abnormal (i.e., HPV-positive or ASC or higher-grade cytology results).


ASC is used to describe “cellular abnormalities that were more marked than those attributable to reactive changes but that fell short of a definitive diagnosis of ‘squamous intraepithelial lesion.’” This interpretation is by far the most common cytologic abnormality, and as a consequence, it precedes the diagnoses of CIN 2-3+ more often than any other cytology result. However, aggressive investigation should be avoided because the ASC diagnosis is poorly reproducible, the risk of cancer is very low (0.1 to 0.2 percent), and the risk of CIN 2-3+ for any individual patient is also low (6.4 to 11.9 percent).

Options for evaluation include immediate colposcopy, triage to colposcopy by HPV DNA testing, or repeat cytologic testing at six and 12 months. Colposcopy provides a rapid diagnosis with the least possible loss to follow-up, but it is expensive and unpleasant for patients. Testing for high-risk HPV types and referral to colposcopy for women who test positive have the advantage of prompt diagnosis and the ability to reassure 44 to 69 percent of patients without colposcopy that their risk of a significant lesion is very low. Reflex testing for HPV if liquid cytology was used, or from a separate sample collected at the time of initial cervical cytologic testing, is preferred for patient convenience and cost-effectiveness. If HPV testing is elected, women whose test results are HPV positive have a 15 to 27 percent chance of having CIN 2-3+ and should be referred for colposcopy. Women who test negative for HPV can be reassured that their risk of having CIN 2-3+ is less than 2 percent, and they can be scheduled for repeat cytologic testing in one year.

The exception to this recommendation for HPV follow-up is the adolescent, for whom the risk of invasive cancer approaches zero and the likelihood of HPV clearance is very high. As an alternative to immediate colposcopy, adolescents with ASC HPV-positive test results may be monitored with cytologic testing at six and 12 months or with a single HPV test at 12 months, with colposcopy for any abnormal cytology result or positive HPV test result.


Low-grade squamous intraepithelial lesion (LSIL) is the second most common abnormal cytology result and is more common in younger populations with larger numbers of recent partners. The risk of CIN 2-3+ at initial colposcopy following an LSIL result is between 15 and 30 percent in most studies. This level of risk is similar to results of initial colposcopy associated with an ASC HPV-positive cytology result in other studies. Therefore, colposcopy is recommended for evaluation of LSIL. For adolescents with LSIL results, it may be reasonable to follow up without immediate colposcopy. LSIL is common in sexually active adolescents because of the recent onset of sexual activity in this group, but clearance of HPV is high and cancer rates are extremely low. Therefore, follow-up recommendations are similar to those for adolescents with ASC HPV-positive results.

CIN 2-3+ has been detected in 24 to 94 percent of patients with cytology results of “ASC—cannot exclude high-grade intraepithelial lesions” (ASC-H). This suggests that colposcopy is an appropriate initial diagnostic intervention. HPV has been detected in 86 percent of women with ASC-H monolayer cytology and in 70 percent of women with ASC-H conventional cytology. Although HPV testing defines a population at low risk, it may not be cost-effective for triage in younger women. In women 30 years and older with ASC-H cytology results, HPV-positive test results decrease dramatically, and triage to colposcopy using HPV testing may be considered. Otherwise, treatment for women with ASC-H should be similar to that for women with LSIL; follow-up of a colposcopy result of CIN 1 or normal should include ▴ cytologic testing at six and 12 months or an HPV DNA test at 12 months, rather than excision.

No dysplastic lesions were identified in nearly one half of women evaluated for ASC-H. This suggests that for women with ASC-H, excision is not warranted in those who have an initial negative colposcopy result. For women with an ASC HPV-positive, ASC-H, or LSIL cytology result and a negative initial colposcopy examination or a histology result of CIN 1, optimal follow-up is repeat cervical cytology screening at six and 12 months or an HPV test at 12 months. If the follow-up cytology result is ASC or higher-grade cytology or a positive HPV test, colposcopy should be repeated.


CIN 2 or CIN 3 has been reported in at least 70 percent of women with cytology results of high-grade squamous intraepithelial lesions (HSIL), and 1 to 2 percent have invasive cancer. Given the level of risk, colposcopy and biopsy of visible lesions are recommended. Endocervical assessment should be performed in nonpregnant patients, and the entire vagina should be examined, especially when a lesion corresponding to the cytology result is not found. An alternative “see and treat” management plan may be used in these patients if a lesion consistent with CIN 2 or CIN 3 is observed. In women who have such a lesion and are not pregnant, loop electrosurgical excision procedure (LEEP) may be performed at the same visit as the colposcopy. In these patients, the cervical biopsy is omitted and an endocervical assessment after the LEEP may be considered.

Because HSIL and ASC-H do not carry the same risk of CIN 2-3+, recommendations for follow-up differ. Excision generally is recommended for women with HSIL cytology results and a negative initial colposcopic evaluation.

Interpretations of HSIL and CIN 2 or CIN 3 are poorly reproducible. Consequently, experts have recommended review of the cytology and histology results in patients with HSIL diagnoses and discrepancies in colposcopic results, although this approach has not been tested in clinical studies. If review is not undertaken or colposcopy results are not satisfactory, excision is recommended. This approach is favored because a single colposcopy can miss CIN 2 or CIN 3, particularly small lesions, and because investigators have documented CIN 2-3+ when examining excision specimens in up to 35 percent of women with HSIL cytology results and negative or noncorrelating (CIN 1) colposcopy results. Adolescents are exceptions to this recommendation because interobserver variability is most pronounced in younger women, the risk of invasive cancer is extremely low, and the likelihood of spontaneous resolution of CIN 1 or CIN 2 is high. Therefore, follow-up with ▴ colposcopy and cytologic testing at four to six months may be undertaken, as long as the colposcopy results are adequate and the endocervical curettage is negative.


Endocervical curettage and colposcopy are both relatively insensitive for AIS and adenocarcinoma, but most women with cytology results of atypical glandular cells (AGC) do not have significant lesions. The initial evaluation of women with AGC results is dictated by the risk of CIN 2-3+, by the possibility that the source of the abnormality may be the endometrium, and by the recognition that the entire endocervix is at risk for AIS. Therefore, colposcopy and endocervical sampling should be included in the initial evaluation of all women with AGC results, except for those with results that specify “atypical endometrial cells.” Women with atypical endometrial cells and a normal endometrial sampling should undergo colposcopy and endocervical sampling. Endometrial sampling is indicated in women with atypical endometrial cells and all women with AGC results who are 35 years and older, as well as those younger than 35 years with abnormal bleeding, morbid obesity, oligomenorrhea, or clinical evaluation suggesting endometrial cancer.

Treatment of women with AGC and negative initial evaluations is determined by the risk that significant disease is present but was not detected. The category “AGC-not otherwise specified” (AGC-NOS) is associated with a low risk of missed disease; follow-up with repeat cytologic testing and endocervical sampling four times at six-month intervals is recommended.

Like squamous CIN, HPV is found in more than 95 percent of AIS and 90 to 100 percent of invasive adenocarcinomas of the cervix. The largest published series of AGC results uniformly evaluated with cervical histology and HPV testing found that 40 of 137 women (29 percent) were HPV positive, including 11 of 12 women with CIN 2 or CIN 3 and all five women with AIS. Similar reports suggest that it is reasonable to monitor women with AGC cytology results, a negative initial evaluation, and a negative HPV test result with a repeat cytology and endocervical sampling in one year rather than requiring four visits at six-month intervals.

For women with results of “AGC–favor neoplasia” or AIS and a negative initial evaluation, or a second AGC-NOS result and a second negative evaluation, the risk of missing a significant lesion is sufficient that excision is warranted. Cold-knife conization is a good choice in this situation because of the prognostic importance in AIS of the pathologic evaluation of margins, which may be obscured by thermal artifact in some LEEP specimens. The rarity of this diagnosis and the difficulty with management may require consultation with a subspecialist.


Untreated CIN 1 confers a risk of 13 percent for diagnosis of CIN 2 or CIN 3 at two-year follow-up, which is the same as the risk for ASC HPV-positive or LSIL cytology results following a negative colposcopy. However, most cases of CIN 1 will remit spontaneously over time. The decision for treatment or observation should be based on the preferences of the patient and the physician. For most women, especially younger women, observation provides the best balance between risk and benefit and should be encouraged. Follow-up of untreated CIN 1 should include two cytology screening tests six months apart, with colposcopy for an ASC or higher-grade result, or a single HPV test at 12 months, with colposcopy if the test result is positive.


CIN 2 and CIN 3 are recognized potential cancer precursors, although CIN 2 is associated with significant spontaneous regression. Evidence suggests that approximately 40 percent of CIN 2 cases regress over two years, whereas regression of CIN 3 is too rare to measure accurately. As a consequence, immediate treatment of CIN 2 and CIN 3 with excision or ablation in nonpregnant patients is recommended. The only exception to this recommendation is that follow-up similar to CIN 1 may be considered in adolescents with CIN 2, whose likelihood of spontaneous clearance is substantial and whose risk of cancer approaches zero. Therefore, care of the adolescent with CIN 2 may be individualized.

In the absence of other indications, hysterectomy is not the initial treatment of choice for patients with CIN 2 or CIN 3. Hysterectomy may be considered for treatment of persistent or recurrent CIN 2 or CIN 3 or when a repeat excision is indicated but technically unfeasible. If excision is indicated, it should be performed (where possible) before hysterectomy to rule out invasive cancer.

Continue Reading

More in AFP

Copyright © 2006 by the American Academy of Family Physicians.

This content is owned by the AAFP. A person viewing it online may make one printout of the material and may use that printout only for his or her personal, non-commercial reference. This material may not otherwise be downloaded, copied, printed, stored, transmitted or reproduced in any medium, whether now known or later invented, except as authorized in writing by the AAFP.  See for copyright questions and/or permission requests.