Cervical Intraepithelial Neoplasia

Cervical screening in pregnancy

Cervical screening aims to reduce the incidence of and mortality from cervical cancer through the detection and effective treatment of premalignant lesions. It is widely accepted that in regions where systematic population-based screening programmes have been established, the incidence and mortality rates of invasive cervical cancer have reduced as a direct consequence, for example, in British Columbia, the UK and the Scandinavian countries .

The UK National Health Service Cervical Screening Programme (NHSCSP) advocates that routine screening is not performed antenatally and should be deferred until after delivery. This is because cytological interpretation of cervical smears collected during pregnancy is difficult due to pregnancy-related changes and treatment for any underlying CIN is not recommended during pregnancy. Cervical sampling may also lead to minor vaginal bleeding, which can cause anxiety to the pregnant woman. If, however, a woman with cytological abnormalities becomes pregnant in the interim, follow-up should not be postponed until after delivery. Under these circumstances, a repeat cervical sample should be taken in the second trimester, or it may be more appropriate to perform colposcopy . Cervical sampling and colposcopy are not associated with adverse pregnancy outcomes such as miscarriage or rupture of membranes .

In countries where an organised screening programme is not available, screening by cytological examination or visual inspection of the cervix with acetic acid may be offered during pregnancy. Although this practice is controversial, supporters identify pregnancy as an ideal time to perform opportunistic screening in women who might not otherwise be screened .

Cervical screening in pregnancy

Cervical screening aims to reduce the incidence of and mortality from cervical cancer through the detection and effective treatment of premalignant lesions. It is widely accepted that in regions where systematic population-based screening programmes have been established, the incidence and mortality rates of invasive cervical cancer have reduced as a direct consequence, for example, in British Columbia, the UK and the Scandinavian countries .

The UK National Health Service Cervical Screening Programme (NHSCSP) advocates that routine screening is not performed antenatally and should be deferred until after delivery. This is because cytological interpretation of cervical smears collected during pregnancy is difficult due to pregnancy-related changes and treatment for any underlying CIN is not recommended during pregnancy. Cervical sampling may also lead to minor vaginal bleeding, which can cause anxiety to the pregnant woman. If, however, a woman with cytological abnormalities becomes pregnant in the interim, follow-up should not be postponed until after delivery. Under these circumstances, a repeat cervical sample should be taken in the second trimester, or it may be more appropriate to perform colposcopy . Cervical sampling and colposcopy are not associated with adverse pregnancy outcomes such as miscarriage or rupture of membranes .

In countries where an organised screening programme is not available, screening by cytological examination or visual inspection of the cervix with acetic acid may be offered during pregnancy. Although this practice is controversial, supporters identify pregnancy as an ideal time to perform opportunistic screening in women who might not otherwise be screened .

Colposcopy in pregnancy

The indications for colposcopy for women with cytological abnormalities during pregnancy are essentially the same as for non-pregnant women. The need for colposcopy during pregnancy may arise as a result of:

• •
  

  Required follow-up of histologically confirmed CIN in women who became pregnant prior to definitive treatment.

  
  
• •
  

  Abnormal cervical cytological results from opportunistic screening at the time of the booking visit.

  
  
• •
  

  Abnormal cervical cytological results in women who have become pregnant while awaiting referral to colposcopy.

It may also arise from the need to exclude early cervical cancer in women who have repeated antepartum haemorrhage in the absence of another obvious cause.

Colposcopy is a safe and effective method for the further evaluation of cytological abnormalities; however, procedural difficulties can arise when performing colposcopy in pregnancy. Firstly, vaginal wall laxity is common, and secondly, the cervix is more vascular during pregnancy leading to increased friability and the possibility of traumatic bleeding .

Colposcopic appearance of CIN

The colposcopic appearances of CIN in pregnancy are generally similar to that of the non-gravid cervix. In non-pregnant women, acetic acid (3 or 5%) is applied to the cervix under direct vision, and the colposcopist needs to differentiate among normal epithelium, low-grade (CIN1) and high-grade (CIN2/3) disease using pattern recognition. A high degree of skill is required to discriminate between changes attributable to the pregnant state and pathological lesions; therefore, these examinations should only be undertaken by an experienced colposcopist.

The cervix undergoes major physiological changes during pregnancy, thereby making interpretation of both cytological and colposcopic findings more difficult. The characteristic changes include an increased cervical volume, stromal oedema, glandular hyperplasia and decidual reaction . There is eversion of the endocervical columnar epithelium on to the ectocervix, which causes active squamous metaplasia, making adequate visualisation of the squamocolumnar junction easier during pregnancy. Conversely, squamous metaplasia can be associated with a mosaic or punctuate surface pattern, making it difficult to distinguish from low-grade CIN . Furthermore, hormonal stimulation of the endocervical glands increases the production of mucus, which can obscure colposcopic visualisation of the cervix .

With regard to cytological analysis, pregnancy-related hormonal stimulation induces changes such as hyperplasia and reactive atypia in squamous and glandular epithelial cells. These changes, combined with the Arias-Stella reaction, make it more difficult to identify atypical glandular cells on cytological specimens . In addition, stromal decidualisation produces cells with large nuclei, which mimic dyskaryotic cells; indeed, to the inexperienced eye, a heavily decidualised cervix can even look malignant. The presence of trophoblastic cells and immature metaplasia in cytological specimens obtained during pregnancy may also be misdiagnosed as moderate or severe dyskaryosis . Despite this, both conventional and liquid-based cytology do demonstrate similar diagnostic features in pregnant and non-pregnant patients .

Follow-up for women who have previously been treated for CIN

If a pregnant woman requires colposcopic or cytological examination after treatment for CIN or follow-up of untreated CIN, her review can safely be postponed until after delivery, unless it is the first clinical review following treatment for CGIN or treatment for CIN2/3 with involved or uncertain margin status. If repeat cytological examination is due and the woman has defaulted from a follow-up appointment prior to pregnancy, consideration should be given to her undergoing cytological or colposcopic examination during pregnancy .

The Natural History of CIN in Pregnancy

A number of cohort studies and retrospective studies have demonstrated the safety of delaying treatment of CIN in pregnant women . The incidence of invasive cervical cancer in pregnancy is relatively low in developing countries affecting approximately 1.2 per 10,000 pregnancies . Pregnancy itself has not been shown to have an adverse impact on the prognosis of invasive cancer .

Paraskevaidis et al. evaluated the safety of cytological and colposcopic surveillance of women with CIN during pregnancy. A total of 98 women with antenatal cytological abnormalities and/or colposcopic diagnosis of CIN were followed up during pregnancy with cytological and colposcopic examination every 2 months. A cytological and colposcopic re-evaluation was carried out 2 months post partum, during which a large-loop excision of the transformation zone (LLETZ) was performed, if clinically indicated. Punch or loop biopsies were only taken during pregnancy if micro-invasion was suspected. In 14 of 39 (35.9%) and in 25 of 52 (48.1%) women with an antenatal impression of CIN1 and CIN2/3, respectively, regression was evident post partum. The investigators concluded that the rate of regression of CIN was high after delivery. The regression rate for CIN1 in this study was double that previously reported for CIN1 in non-pregnant women . These findings are consistent with other studies, which have demonstrated that the risk of progression of CIN during pregnancy is low. Coppola et al. reviewed 34 women with cytological abnormalities during pregnancy over a 5-year period, of whom 26 underwent post-partum assessment. One woman was treated with a cone biopsy because of suspected microinvasive disease in pregnancy. Of the remaining 25 women, two (8%) had progressive disease post partum and 23 (92%) had either persistent or regressive disease. In a larger study involving 153 women who underwent colposcopic biopsies during pregnancy, 7% of the patients with CIN2 were found to have CIN3 at the time of post-partum evaluation . Spontaneous post-partum regression was reported in up to 70% of histologically confirmed CIN2/3 lesions. It should be noted that the so-called progression of CIN during pregnancy could be simply due to the limitations of a punch biopsy when compared with an excisional biopsy.

It has been suggested that the regression of CIN during pregnancy may be attributable to cervical trauma from birth or transient ischaemic changes in the cervical epithelium and the subsequent activation of inflammation and repair mechanisms, during cervical ripening and vaginal delivery . Whether or not the mode of delivery plays a role in the regression of CIN remains a matter of debate . Ahdoot et al. reported spontaneous regression in 60% of women with CIN2/3 who had a vaginal delivery, but none of the patients who delivered by caesarean section demonstrated regression.

Management of cervical cytological abnormalities in pregnant women

The principal aim of colposcopy for pregnant women is to exclude invasive cancer and to defer biopsy and definitive treatment until the post-partum period . During pregnancy, it is acceptable to make a provisional diagnosis of CIN based only on the cytological and colposcopic findings without histological confirmation. The European Cervical Cancer Screening Network and NHSCSP guidelines recommend that women with low-grade cytological abnormalities, and in whom colposcopic assessment does not suggest high-grade disease, should undergo repeat colposcopy 3–4 months post partum. Women with high-grade cytological abnormalities and in whom colposcopy has excluded invasive disease should be reviewed at 3-month intervals during pregnancy, with a final assessment 3–4 months following delivery ( Fig.1 ). A decision should be made on whether treatment is required during postnatal review.

Algorithm for the management of cytological abnormalities in pregnancy.

Although several studies have established the safety of colposcopically directed biopsies, the clinical value of such biopsies in pregnancy remains unclear. A retrospective study by Economos et al. compared cytological, colposcopic and histological diagnoses in 112 women with cytological abnormalities in pregnancy. The investigators found 91% concordance within one degree of severity when comparing cytological findings with the colposcopic impression, and 95% concordance within one degree of severity when comparing colposcopic and histological diagnosis. No cases of invasive disease were overlooked and complications of biopsies were minimal. Essentially, it was concluded that directed biopsies were safe and reliable in women who were referred with abnormal smears in pregnancy. Other studies have found similar results . However, the standard practice in the UK and continental Europe is to reserve cervical biopsies for cases where invasive disease is suspected, either clinically or colposcopically .

When invasive disease is suspected, conventional punch biopsies may not be adequate as they cannot safely confirm or exclude micro-invasion, and they may fail to give a representative sample. Multiple punch biopsies are also not advisable as they may cause trauma and significant bleeding . Consequently, a wedge biopsy using a small loop, needle diathermy or a knife, sampling the worst affected area of the epithelium identified colposcopically, is a better option. As there is a significant risk of haemorrhage with these biopsies, they are usually carried out in the operating theatre. Decisions regarding the need for a wedge biopsy and the procedure itself should be performed by an experienced clinician with relevant expertise.

The guidelines of the American Society of Colposcopy and Cervical Pathology (ASCCP) differ from the European practice. Firstly, colposcopy is recommended for non-adolescent pregnant women with low-grade cytological abnormalities such as low-grade squamous intraepithelial lesion (LSIL), but unlike in Europe where colposcopic inspection is recommended antenatally, clinicians in the US have the option of deferring colposcopy until at least 6 weeks post partum. If women undergo an initial colposcopy antenatally, provided there is no evidence of CIN2/3, a post-partum follow-up is recommended. The ASCCP advises pregnant women with high-grade cytological abnormalities such as high-grade squamous epithelial lesions (HSIL) to undergo colposcopy. While the European consensus recommends deferring biopsies until after delivery unless invasive disease is suspected, the ASCCP suggests biopsy of lesions suspicious of CIN2/3 or invasive disease. If CIN2/3 is confirmed by histology, then colposcopic and repeat cytological examination and are recommended every 3 months during pregnancy, but repeat biopsies are not advisable unless colposcopic or cytological assessment suggests progression to invasive disease. Conversely, women in whom CIN2/3 is excluded should await follow-up a minimum of 6 weeks after delivery.

Excisional biopsy and treatment in pregnancy

As highlighted earlier, excisional biopsy should only be performed in pregnancy in order to exclude invasive cancer. Treatment for preinvasive lesions during pregnancy should be avoided, as it is associated with an increased risk of significant morbidity, and there is also an excessively high risk of incomplete excision and disease persistence . Reports of treatment of CIN in pregnancy are influenced by case selection and expertise, both of which affect treatment outcomes. These factors should be borne in mind when comparing reports.

Robinson et al. reported on the outcomes of 20 women who underwent LLETZ for CIN between 8 and 34 weeks of gestation. The investigators found a significant complication rate of 25%. Two patients required blood transfusion and three others had a preterm delivery. Of note, all five women with a significant complication underwent the procedure between 27 and 34 weeks. Only 70% of the histological specimens actually contained CIN and 57% had involved margins, with residual disease noted in nine women. The initial LLETZ specimen of three women with disease at follow-up did not contain CIN. In a smaller study by Matsuhashi et al. (2000), nine patients underwent an LLETZ under a local anaesthetic within the first 14 weeks of pregnancy. In contrast to the findings of Robinson et al. , Matsuhashi et al. noted no significant intraoperative or post-operative complications, and all patients delivered at term. In the latter study, five women had a histological diagnosis of CIN3 and two had microinvasive disease. Margins were incompletely excised in two women, who subsequently underwent a repeat LLETZ post partum.

Glandular lesions in pregnancy

Adenocarcinoma in situ and primary adenocarcinoma of the cervix are uncommon, accounting for less than 15% of all cervical cancers. Data regarding the significance of glandular atypia during pregnancy are limited. A study by Chhieng et al. evaluated 35 women with atypical glandular cells of uncertain origin on cytological assessment. Twenty-seven women underwent follow-up: 17 underwent colposcopy and directed biopsy and 10 repeat cytological examination. Almost one-third had concomitant squamous intraepithelial neoplasia, on subsequent biopsy, while the remainder had benign histology, which included chronic cervicitis, endocervical polyps, microglandular hyperplasia and Arias-Stella reaction. None of the women referred with atypical glandular cells actually had evidence of a significant glandular lesion. The natural history and implications of glandular atypia in pregnancy are poorly understood. In non-pregnant women, glandular atypia is often associated with squamous lesions; therefore, these women require a close follow-up with colposcopy during pregnancy and post-partum directed biopsy of suspicious areas .