Normal colposcopic findings following LBC prediction of Negative, LSIL or HSIL

Systematic reviews were conducted to answer the following questions:

• For women with a positive oncogenic HPV test result (not in treatment follow-up) with negative or LSIL cytology and normal colposcopy, what is the safety and effectiveness of 12-month follow-up testing with a HPV test alone, compared with co-testing?
• For women with a positive oncogenic HPV test result (not in treatment follow-up) with negative or LSIL cytology and normal colposcopy, which factors predict the presence of high-grade cervical neoplastic disease (CIN2, CIN3, cervical cancer, adenocarcinoma in situ (AIS) or cervical cancer mortality)?
• For women with a positive oncogenic HPV test result and pHSIL/HSIL referral cytology but pLSIL/LSIL or less after cytologic review, and colposcopy is normal (negative), what is the safety and effectiveness of conservative management compared with excision of the TZ?
• For women with a positive oncogenic HPV test result and pHSIL/HSIL referral cytology, confirmed pHSIL/HSIL after cytologic review, and colposcopy is normal (negative), what is the safety and effectiveness of cytological and colposcopic follow-up at 3–6 months, compared with excision of the TZ?

Outcomes for women with normal colposcopic findings following referral cytology prediction of Negative or LSIL

No randomised or pseudorandomised controlled trials were identified that compared HPV testing with co-testing (HPV and LBC) at 12-month follow-up for women with a positive oncogenic HPV (any type) test result, a cytology report of negative or LSIL and normal colposcopy, and which reported high grade-disease outcomes.

No studies were found that reported the risk of high-grade disease associated with follow-up cytology or HPV status.

Four studies reported the risks associated with referral cytology and/or HPV status for women reported as pLSIL/LSIL and normal (negative) colposcopic findings: two prospective cohort studies^[1][2] (level II evidence) and two retrospective cohort studies^[3][4] (level III-2 evidence) were identified. Follow-up ranged from 1 to 3 years. All four studies were considered at high risk of bias; none of the studies were specifically designed to answer the PICO question, so it was not clear as to whether women with different baseline cytology results or HPV status were similar in terms of important confounders such as smoking status. Furthermore, important study design aspects such as the potential blinded reading of subsequent colposcopies and histopathology (with respect to the baseline test status) were not described.

Two of the studies^[3][4] examined the risks of cervical intraepithelial neoplasia (CIN) grade 3 or higher (CIN3+) associated with different baseline cytology results, and one study^[2] examined the risks of CIN3+ associated with baseline HPV-positive ASC-US and LSIL (regardless of HPV status).Three studies^[3][4][1] examined the risks of CIN2+ disease associated with different baseline cytology results, of which one study^[1] examined the risks of CIN2+ associated with different baseline HPV status. Two studies^[1][4] reported the risks associated with baseline cytology results in women who were HPV positive.

One of these studies[3] did not report HPV status. The remaining studies provided the following evidence:

• In a cohort of women with a positive oncogenic HPV (any type) test result with normal colposcopy after referral cytology reported as ‘borderline dyskaryosis’ or ‘mild dyskaryosis’, rates of later detection of CIN3+ (median follow-up 27 months) were 3.5% and 2.1%, respectively.^[4] This cohort may not be representative of women with a LBC prediction of pLSIL/LSIL identified after primary HPV testing in the renewed NCSP; the HPV test in this UK study used a higher cut-off than recommended, and the ‘borderline dyskaryosis’ group could have included women with pHSIL.
• In a cohort of women with a referral cytology report of LSIL and normal colposcopy, the rate of CIN3+ after 1 year follow-up was 4.0%.^[2] This cohort included women who tested HPV positive and women who tested HPV negative.
• In a cohort of women with a positive oncogenic HPV (any type) test result with normal colposcopy after referral cytology reported as ‘borderline dyskaryosis’ or ‘mild dyskaryosis’, rates of later detection of CIN2+ (median follow-up 2.6 years) were 6.2% and 12.9%, respectively.^[1]

The systematic reviews and their findings are described in detail in the Technical report.

As no studies were found that reported the risks of high-grade disease associated with follow-up cytology or HPV status, there was no directly relevant evidence on which to base an evidence-based recommendation. Detection of HPV, especially persistent HPV 16/18, is associated with an increased risk of high grade cervical lesions, and the HPV test is more sensitive than cytology (see Chapter 2. The rationale for primary HPV screening).

Outcomes for women with normal colposcopy following referral cytology prediction of HSIL

Systematic literature searches did not identify any studies directly addressing the management of women with HSIL cytology and a normal (negative) colposcopy. The search strategies and inclusion and exclusion criteria used are described in detail in the Technical report. .

In the absence of any direct evidence from the systematic review, a general review of the literature was performed on the management of women with cytological prediction of pHSIL/HSIL and normal (negative) colposcopy was undertaken to inform relevant consensus-based recommendations. No studies were found that reported outcomes for women followed up after referral cytology prediction of pHSIL/HSIL and normal colposcopic findings, and which reported the results of cytopathology review.

One cross-sectional cohort study^[5] reported outcomes for women participating in conventional cytology (Pap test) screening with no history of a cytological abnormality, who had a cytological prediction of HSIL between 2000 and 2007. Of 340 women who underwent colposcopy, 17 had normal colposcopic findings. Biopsy was performed in nine of these women, including endocervical curettage in at least four women.^[5] Ages of the women and results of cytopathology review (if performed) were not reported.

Of the 17 women with normal colposcopy (HPV status unknown), two (11.8%) were diagnosed with cervical adenocarcinoma and another two (11.8%) with AIS.^[5] Findings for the other 13 women were not reported. No other cases of cervical adenocarcinoma or cervical adenocarcinoma in situ were identified among the 331 women with HSIL who underwent biopsy.^[5] The degree to which these findings can be generalised to women in the renewed NCSP is limited, because HPV status and the findings of cytopathology review are not available.

Another retrospective cohort study^[6] reported outcomes for a subgroup of 59 women (mean age 26.8 years) who underwent a loop electrosurgical excision procedure (LEEP) after a cytological prediction of HSIL and a colposcopically directed target biopsy finding of no abnormality detected (n = 34) or CIN1 (n = 25). On excisional biopsy, histologically confirmed CIN3 was diagnosed in 14 (41%) women with normal target biopsy and 16 (64%) women with CIN1 on target biopsy. The degree to which these findings can be generalised to women in the renewed NCSP is limited, because HPV status was unknown and the study did not report whether HSIL was reported at initial referral cytology or on cytopathology review. However, this study did demonstrate the failure of colposcopically directed biopsy to detect the high-grade lesion in 41% of cases, as high grade abnormality was subsequently confirmed in an excision biopsy.^[6]

A summary of the literature considered can be found in the Technical report.

Overall, this body of evidence is insufficient to enable accurate prediction of risk in women with normal colposcopic findings despite a LBC report of HSIL.

In Australia, it is recommended that discordant findings between referral cytology and colposcopy warrant review of the cytology prior to further management decisions.

Current evidence does suggest that, in women with a normal colposcopy and HSIL on cytopathology review, the risk of CIN3 and/or invasive cervical cancer is high enough to warrant diagnostic excision of the TZ. In women with a normal colposcopy and pHSIL on cytopathology review, the risk of CIN3 and/or invasive cervical cancer is high enough to warrant diagnostic excision of the transformation zone in most cases, but in some situations a short period of observation may be appropriate (see Chapter 9. Management of histologically confirmed low-grade squamous abnormalities).

Flowchart 8.1. Normal colposcopy after LBC prediction of Negative, pLSIL or LSIL, following detection of HPV (any type)

Flowchart 8.2. Normal colposcopy after LBC prediction of possible HSIL

Flowchart 8.3. Normal colposcopy after LBC prediction of HSIL

If these recommendations, including review of cytology, are followed for women who have normal colposcopy in the presence of referral LBC predicting low-grade or high-grade grade lesions, they will benefit by avoiding over-investigation or unnecessary treatment.

See Chapter 5. Benefits, harms and cost-effectiveness of cervical screening in the renewed National Cervical Screening Program (NCSP).

Clinical practice

It is not anticipated that there will be a significant change to clinical practice, apart from the addition of HPV testing to the recommended surveillance.

Resourcing

No material changes to the costs are anticipated.