SEARCH

SEARCH BY CITATION

Keywords:

  • Cytology;
  • follow up;
  • HPV testing

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Details of ethics approval
  9. Contribution to authorship
  10. References

Objective  To evaluate human papillomavirus (HPV) testing in combination with cytology in the follow up of treated women.

Design  A prospective study.

Setting  Three UK centres: Manchester, Aberdeen and London.

Population or sample  Women treated for cervical intraepithelial neoplasia (CIN).

Methods  Women were recruited at 6 months of follow up, and cytology and HPV testing was carried out at 6 and 12 months. If either or both results were positive, colposcopy and if appropriate, a biopsy and retreatment was performed. At 24 months, cytology alone was performed.

Main outcome measures  Cytology and histology at 6, 12 and 24 months.

Results  Nine hundred and seventeen women were recruited at 6 months of follow up, with 778 (85%) and 707 (77.1%) being recruited at 12 and 24 months, respectively. At recruitment, 700 women had had high-grade CIN (grades 2 or 3) and 217 had CIN1. At 6 months, 14.6% were HPV positive and 10.7% had non-negative cytology. Of those with negative cytology, 9% were HPV positive. Of the 744 women who were cytology negative/HPV negative at baseline, 3 women with CIN2, 1 with CIN3, 1 with cancer and 1 with vaginal intraepithelial neoplasia (VAIN)1 were identified at 24 months. Nine of 10 cases of CIN3/cervical glandular intraepithelial neoplasia (CGIN) occurred in HPV-positive women. At 23 months, cancer was identified in a woman treated for CGIN with clear resection margins, who had been cytology negative/HPV negative at both 6 and 12 months.

Conclusions  Women who are cytology negative and HPV negative at 6 months after treatment for CIN can safely be returned to 3-year recall.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Details of ethics approval
  9. Contribution to authorship
  10. References

Women treated for cervical intraepithelial neoplasia (CIN) by conservative methods remain at increased risk of developing CIN and cancer for up to 10 years and more when compared with unaffected, screened women.1 This has led to more intensive follow up, the national guideline currently being annual cytology for 2 years after treatment of CIN1 and for 10 years for CIN2/3.2 The policy of post-treatment cytological surveillance means that treatment failure should be detected quickly and repeat conservative treatment is usually possible. This has reduced the risk of post-treatment cancer to around 1:200 over an 8-year period.1 The majority of treatment failures are due to residual disease and are detected cytologically within 2 years.3 Lesions discovered after 3 years are generally regarded as reoccurrence of CIN.

It has been known for a number of years that persistent infection by so-called high-risk human papillomavirus (HPV) is a necessary event in cervical carcinogenesis and virtually 100% of cervical cancers, either squamous or glandular, contain HPV DNA. In CIN3, the precursor lesion of cancer, over 90% of lesions are HPV positive. These observations explain why HPV has a very high negative predictive value and high sensitivity either as a screening test or as a triage of mild cytological abnormalities. In terms of follow up, this negative predictive value is also of potential clinical utility, and a number of studies have confirmed its sensitivity to detect CIN after treatment.4 This means that women who are HPV negative after treatment are at very low risk of having residual disease or developing recurrent CIN within 3–5 years. In a case–control study matched for excision margins, it was shown that the presence of high-risk HPV was the strongest risk factor for treatment failure.5

This study was designed to demonstrate the clinical utility of HPV testing as an adjuvant to cytology by following a large group of treated women. The objective was to confirm that women who were both cytology negative and HPV negative at 6 months could be shown to be at significantly low risk of developing CIN2/3 over a 2-year period and safe to return to routine recall at 3 or 5 years.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Details of ethics approval
  9. Contribution to authorship
  10. References

Women treated for CIN at St Mary’s Hospital, Manchester, Royal Free Hospital, London, and Aberdeen Royal Infirmary were seen routinely at the colposcopy clinic 6 months after treatment and provided with an information sheet informing them of the HPV test as part of the follow-up protocol.

All women underwent cytology and HPV testing at 6 and 12 months and cytology alone at 24 months. We did not test for HPV at 24 months because we expected the majority of treatment failure would have been identified by 12 months of follow up and mandating a further colposcopy for a HPV-positive test would not be justifiable with negative cytology. In Aberdeen and Manchester, the cytology was liquid based, and in the Royal Free, conventional cytology was performed. All HPV testing was performed using the Hybrid Capture 2 (HC2; Digene, Gaithersburg, MD, USA) test according to the manufacturer’s instructions at a positive cutoff of 1 RLU/coequivalent to 1 pg/ml. No HPV typing was performed. Follow-up colposcopy as a routine was used according to local practice but was performed at 6 and 12 months if either the cytology was abnormal or HPV test was positive, together with a biopsy if colposcopy was abnormal.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Details of ethics approval
  9. Contribution to authorship
  10. References

The flow of women through the follow-up protocol is shown in Figure 1. Nine hundred and seventeen women with a median age of 31.5 years (range 15–72 years) were seen at 6 months after treatment of CIN. CIN3 had been present in 365 women, CIN2 in 326 women, CIN1 in 217 women and cervical glandular intraepithelial neoplasia (CGIN) in 9 women. Eight hundred and twenty-seven women had been treated by large loop excision of the transformation zone, 81 by laser cone, 2 by cold knife cone and 7 had missing data. In the 835 (91%) women with reported excision margins, 589 (70.5%) had clear margins.

image

Figure 1. Flow of women through study protocol.

Download figure to PowerPoint

Baseline follow-up data showing age (less than 35 years and more than or equal to 35 years) and excision margin by HPV and cytology status are shown in Table 1. Of the 917 treated women, 98 (10.7%) had non-negative cytology including 86 (9.4%) borderline or mild dyskaryosis and 8 (0.87%) had moderate or severe dyskaryosis. One hundred and thirty-four (14.6%) women were HPV positive. Of the 783 HPV-negative women, 39 (4.9%) had non-negative cytology compared with 59/134 (44.0%) HPV-positive women. There were 744 (81.1%) women who were cytology negative and HPV negative. Among the 589 women whose lesions had clear margins, 75 (12.7%) were HPV positive compared with 43/246 (17.5%) women with equivocal or involved margins. The non-negative cytology rates in women with clear and equivocal/involved margins were lower than the HPV-positive rates: 52/589 (8.8%) and 34/246 (13.4%), respectively. The total number of nine CIN2+ in the incomplete excision group accounts for 34.6% of all CIN2+, and this group accounted for 27% of the whole study cohort. In 82 (9%) women, the margin status was missing or unknown; it is not known in what proportion exactly, the pathologist had not reported it.

Table 1.  Age and excision status of treated women by HPV status and cytology at 6 months follow up
 HPV negativeHPV positiveTotal (n = 917)
Cytology negative (n = 744), n (%)Cytology ≥ borderline (n = 39), n (%)Cytology negative (n = 75), n (%)Cytology ≥ borderline (n = 59), n (%)
Age (years)
<35523 (79.0)29 (4.4)66 (10)44 (6.6)662 (100)
≥35209 (87.1)10 (4.2)9 (3.7)12 (5)240 (100)
Missing12003 
Excision margins involved
Involved188 (76.4)15 (6.1)24 (9.8)19 (7.7)246 (100)
Clear491 (83.4)23 (3.9)46 (7.8)29 (4.9)589 (100)
Unknown/missing651511 
Grade of disease
CIN1177 (81.5)12 (5.5)13 (6)15 (7)217 (100)
CIN2/3+567 (81.0)27 (3.8)62 (8.9)44 (6.3)700 (100)

The grades of cytology lesions confirmed at 6 months (baseline) of follow up are shown in Table 2. Twelve women were retreated: 5 with CIN3, 1 with CGIN, 4 with CIN2 and 2/13 women with CIN1. Cytological outcomes at 12 and 24 months of follow up are shown in Table 3 according to baseline HPV and cytology status.

Table 2.  Grades of cytology at 6 (baseline), 12 and 24 months of follow up according to baseline cytology and HPV status
 HPV negativeHPV positiveTotal
Cytology negative (n = 744)Cytology ≥ borderline (n = 39)Cytology negative (n = 75)Cytology ≥ borderline (n = 59)
6 months
Negative74475819
Borderline302757
Mild52429
Moderate33
Severe55
Inadequate44
Defaulted0
12 months
Negative620274920716
Borderline14581542
Mild213814
Moderate
Severe2 13
Inadequate112
Defaulted10551416140
24 months
Negative545245332654
Borderline1533627
Mild124714
Moderate11226
Severe213
Inadequate112
Adenocarcinoma11
Defaulted17971212210
Table 3.  Grades of histopathology of detected lesions at 6 (baseline), 12 and 24 months of follow up according to baseline cytology and HPV status
 HPV negativeHPV positive
Cytology negative (n = 744)Cytology ≥ borderline (n = 39)Cytology negative (n = 75)Cytology ≥ borderline (n = 59)
6 months
CIN18122
CIN2121
CIN35
CGIN1
Cancer
VAIN2/32
12 months
CIN1111
CIN221
CIN312
Cancer
24 months
CIN1112
CIN2212
CIN31
Cancer1
VAIN11 

HPV status at 12 months according to HPV and cytology status at baseline is shown in Table 4. At 12 months, 778 (85%) women were seen and of the 660 who had HPV testing, 76 (11.5%) were positive, of whom 27 (4.1%) were HPV negative at baseline follow up, suggesting that a reinfection rate of 27/660 (4.1%) over 6 months. Of the 54 cytology negative/HPV-positive women at 6 months after treatment, 31 persisted as HPV positive by HC2 at 12 months and 23 had become HPV negative by HC2. There were no typing data to verify whether or not these were true persisting infection.

Table 4.  HPV status at 12 months of follow up according to HPV and cytology status at baseline
HPV status at 12 months post-treatmentHPV negative at baselineHPV positive at baselineTotal
Cytology negative (n = 744)Cytology ≥ borderline (n = 39)Cytology negative (n = 75)Cytology ≥ borderline (n = 59)
Negative518262317584
Positive234311876
Not performed/defaulted20392124257

At 12 months, 778 women were seen. Cytology was performed on 777 women and 660 had a HPV test. Three (0.4%) women had severe dyskaryosis, none had moderate dyskaryosis, 14 (1.8%) had mild dyskaryosis and 42 (5.4%) had borderline dyskaryosis. CIN3 was detected in three women, all of whom tested HPV positive at baseline. CIN2 lesions were detected in three women, two of whom tested HPV negative at baseline.

At 24 months, 707 (77.1%) women were seen including 58 women who had not attended at 12 months. Three (0.4%) had severe dyskaryosis, 6 (0.8%) had moderate dyskaryosis, 14 (2%) had mild dyskaryosis and 27 (3.8%) had borderline dyskaryosis. Among the 58 women who had defaulted at 12 months, only one CIN1 lesion was identified. One new CIN3 lesion and four CIN2 lesions were detected, of which the CIN3 lesion was cytology negative and HPV negative at baseline. One woman who was cytology negative/HPV negative at baseline was found to have a stage III adenocarcinoma. This woman originally had CGIN with clear margins. Subsequent review confirmed an absence of invasive disease in the original biopsy. One possibility is that she had microinvasive adenocarcinoma higher in the cervical canal undetected by the follow-up tests.

In total, the detected high-grade CIN rate was 12/917 (1.3%) at 6 months, 6/778 (0.77%) at 12 months and 6/707 (0.85%) at 24 months. Of those 24 lesions, just one CIN3 and five CIN2 lesions were HPV negative at baseline.

Out of 744 cytology negative/HPV-negative women at baseline, 302 underwent routine colposcopy as part of local protocol. In all but 25 women, colposcopy revealed no abnormality. In 23 of the 25 women, a biopsy was taken which revealed 8 women with CIN1, 1 with CIN2 and 15 with no significant abnormality. The outcome of 744 women who were cytology negative/HPV negative at baseline is shown in Table 5 for the 640 who attended at 12 months; 104 defaulted. Of the 640 women who attended at 12 months, 505 (79%) remained cytology negative/HPV negative, 18 had developed a cytological abnormality and 18 (2.8%) remained cytology negative but became HPV positive. Two CIN2 lesions were detected at 12 months. At 24 months, 405 of the 505 cytology negative/HPV negative remained so. There were nine borderline dyskaryosis, one mild dyskaryosis and one adenocarcinoma with one vaginal intraepithelial neoplasia (VAIN)1 and one adenocarcinoma detected histologically. Eighty-nine women defaulted.

Table 5.  Cytology and HPV results at 12 months for 640 women who were cytology negative and HPV negative at baseline
SmearHPV
PositiveNegativeMissing
  • *

    Defaulted.

Negative1850597
Borderline392
Mild2
Severe2
Inadequate1
Missing1104*

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Details of ethics approval
  9. Contribution to authorship
  10. References

The purpose of follow up after treatment of CIN is to detect residual disease in the short term, and new disease after follow up for at least 3–5 years. The overall rate of treatment failure in this group of women over a 2-year period was 42/947 (4.3%), which is very similar to reported rates in another UK study.1

The rationale of adding of HPV testing to cytology is designed to increase the sensitivity of detection of HPV-positive disease. In this series, there were 77 women who were cytology negative and HPV positive at baseline follow up, in whom four CIN2 and one CIN3 lesions were detected over 2 years. This represented more high-grade CIN than was detected among the ten times larger cytology negative/HPV-negative group. It is clear that HPV testing can add immediate sensitivity by detecting high-grade CIN not detected cytologically at 6 months, although such lesions could well be detected in subsequent annual follow-up cytology. In terms of follow up, the real clinical utility of HPV testing is its negative predictive value. In those women who were HPV negative at the 6-month baseline follow up, the risk of residual CIN3 is negligible. The advanced adenocarcinoma found at 2 years had been preceded by negative tests at 6 and 12 months. Occasional cases such as this will occur with more prolonged follow up. In the large UK series often referred to as the benchmark of treatment failure rates, there was an incidence of invasive disease after treatment of 5.8 per 100 women over 8 years and the risk did not change significantly during that time.1

These data indicate that the cumulative risk of residual disease is so low after 24 months of follow up that a woman who is cytology negative/HPV negative at 6 months can safely be recalled at 3 years rather than undergo 10 years of annual follow up as currently recommended in national guidelines. Further evidence of the safety of returning these women to routine recall lies in the contemporaneous colposcopy findings, which detected only one CIN2 (and eight CIN1) lesion among 313 such women.

What should be the response to women who are HPV positive and/or have an abnormal cytology result 6 months following treatment? This accounted for 19% of the treated women compared with 14.6% who were HPV positive and it would seem reasonable to offer colposcopy to all these women to detect any underlying disease. The added value of cytology over HPV testing is, however, debatable; among the small group of 39 HPV-negative women with non-negative cytology, only two CIN2 lesions were detected by 24 months and would have been very unlikely to have progressed to invasive disease within the routine screening interval. A recent publication6 has indicated a relative risk of 5.5 for post-treatment disease incomplete excision compared with complete excision. Despite this, the reliability of HPV negative/cytology negative in terms of negative predictive value was similar for the incomplete excision and complete excision groups, which confirms that HPV status is more dominant than margin status in terms of risk.

We would recommend that for the time being the follow-up protocol would include both HPV testing and cytology, but longer term, evaluation with larger cohorts of treated women may support reserving cytology as a reflex test for HPV-positive women.

This large prospective study provides further strong evidence to support the use of HPV testing in follow up to reduce the number of women requiring prolonged follow up and at the same time increase detection of residual disease. Over 80% of women could be safely returned to 3- or 5-yearly recall after 6 months, with considerably greater convenience and less anxiety for women. For those with a positive HPV test and/or abnormal cytology, a reflex colposcopy would help accelerate the diagnosis of residual disease. For those women in whom colposcopy is negative, a second round of HPV testing and cytology at 12 months will identify most remaining treatment failures.

The net effect of restricting follow up for HPV-negative women would certainly be fewer cytology samples, although the impact on total colposcopy provision is less certain. Both human and financial resources would thus be better targeted at the at-risk population who would be better protected from disease, and women who were not at risk could be safely reassured, which would be of psychological benefit to them and represent more efficient use of NHS resources. Further planned confirmation of the acceptability and effectiveness of HPV testing in sentinel sites should lead to full implementation in the NHS Cervical Screening Programme.

Details of ethics approval

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Details of ethics approval
  9. Contribution to authorship
  10. References

Ethical approval was not required as this was an implementation pilot within the NHS Cervical Screening Programme.

Contribution to authorship

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Details of ethics approval
  9. Contribution to authorship
  10. References

H.C.K. and P.G.W. contributed to conception and design of the protocol, drafting the article and final approval of the article. L.N. contributed to analysis and interpretation of data, revision of the article and final approval of the article. R.H. contributed to the acquisition of data, revision of the article and final approval of the article. J.P. contributed to conception and design of the protocol, revision of the article and final approval of the article. G.B.A., A.S., J.W. and C.M. contributed to the acquisition of data, revision of the article and final approval of the article. M.E.C. contributed to conception and design of the protocol, drafting the article and final approval of the article.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Details of ethics approval
  9. Contribution to authorship
  10. References
  • 1
    Soutter WP, De Barros Lopes A, Fletcher A, Monaghan JM, Duncan ID, Paraskevaidis E, et al. Invasive cervical cancer after conservative therapy for cervical intraepithelial neoplasia. Lancet 1997;349:97880.
  • 2
    NHSCSP. Colposcopy and Programme Management: Guidelines for the NHS Cervical Screening Programme. NHSCSP publication number 20. Sheffield, UK: NHS Cancer Screening Programmes, 2004.
  • 3
    Paraskevaidis E, Jandial L, Mann EM, Fisher PM, Kitchener HC. Pattern of treatment failure following laser for cervical intraepithelial neoplasia: implications for follow-up protocol. Obstet Gynecol 1991;78:803.
  • 4
    Arbyn M, Paraskevaidis E, Martin-Hirsch P, Prendiville W, Dillner J. Clinical utility of HPV-DNA detection: triage of minor cervical lesions, follow-up of women treated for high-grade CIN: an update of pooled evidence. Gynecol Oncol 2005;99 (3 Suppl 1):S711.
  • 5
    Acladious NN, Sutton C, Mandal D, Hopkins R, Zaklama M, Kitchener H. Persistent human papillomavirus infection and smoking increase risk of failure of treatment of cervical intraepithelial neoplasia (CIN). Int J Cancer 2002;98:4359.
  • 6
    Ghaem-Maghami S, Sagi S, Majeed G, Soutter WP. Incomplete excision of cervical intraepithelial neoplasia and risk of treatment failure: a meta-analysis. Lancet Oncol 2007;8:98593.