The management of women with abnormal cytology in pregnancy represents both a diagnostic and a therapeutic challenge for colposcopists. The emphasis should be on diagnosis and confirmation of cervical precancer (Cervical intraepithelial neoplasia (CIN) or Adenocarcinoma in situ (AIS), thus excluding invasive cancer). Following an initial assessment, careful follow-up is essential. This must include colposcopy and take into account the physiological changes of the cervix during pregnancy and the puerperium. The management of women with invasive cancer diagnosed during pregnancy depends on the gestation at diagnosis and requires careful assessment and multidisciplinary planning.
Cervical cytology screening programmes aim to reduce both the incidence of and mortality from cervical cancer by the detection and effective treatment of pre-invasive lesions. In countries where these programmes have been well organized, especially in Scandinavia and British Columbia, a significant reduction in both the mortality and incidence of cervical cancer has resulted. More recently, improvements in the British NHS Cervical Screening Programme have resulted in a significant reduction in the incidence of invasive cancer of the cervix from 16 per 100 000 in 1986 to 9.3 per 100 000 in 1997. A similar reduction in the mortality of cervical cancer has resulted with the rate falling by 7% per year.
Cervical screening in pregnancy is controversial. This practice is more common in countries without organised cervical screening programmes. Advocates for routine screening in pregnancy regard this as an ideal opportunity to perform a test that might not otherwise be done. The relative immunosuppressed state of pregnancy and the relative frequency of cases of clinically obvious genital and vaginal warts and, by extension, infection with the human papillomavirus (HPV) have been cited as evidence that this is a time which is suitable for the detection of previously unrecognised cervical precancerous abnormalities. On the other hand, cervical smears taken during the antenatal period are more likely to be unsatisfactory and can detect cytologically normal variants, which can cause unnecessary intervention, with the resultant increase in maternal anxiety. The yield of cervical abnormalities in this population of women is a marker of their screening history – with increased numbers of abnormalities detected in populations with low coverage of screening, a factor regarded as important, particularly with the increasing age at first pregnancy. One of the difficulties with this approach is that women can associate screening with pregnancy and stop having smear tests when they stop having children. In countries with organised cervical screening programmes, the case for interval cervical smear tests in well-screened women has not been proven and policymakers make it clear that only those women with an inadequate smear history should be screened in pregnancy ( Fig. 1 ).
In general, the management of women with abnormal smear has radically changed with the introduction of colposcopy in the 1970s as a diagnostic technique for identifying the probable site of the cytological abnormalities. This is a subjective assessment, which relies on pattern recognition, to discriminate between normal and abnormal tissue as well as between grades of abnormality and, used on its own, is not an exact science with the possibility of underestimating and overestimating the degree of cervical intraepithelial neoplasia (CIN). It is, however, generally an effective means of obtaining a confirmatory biopsy and directing treatment. The introduction of colposcopy has radically altered the management of pregnant women with abnormal smears with the adoption of conservative management with conisation during pregnancy reserved for those with suspected invasive cancer. Because therapy for pre-invasive disease can safely be postponed until the postpartum period, the ability to distinguish CIN from invasive cancer has been a major step forward in the management of these women.
Changes to the cervix in pregnancy and the puerperium
Major physiological changes to the uterine cervix are apparent from the first few weeks of pregnancy. These changes can alter the colposcopic appearance and need to be borne in mind when performing colposcopy on pregnant women. Hypertrophy of the stroma of the cervix results in increased eversion of the cervix with appearance of an exaggerated ectropion. This can displace the transformation zone (TZ) laterally. The colposcopist should bear this in mind and remember to evaluate the vaginal fornices to accurately diagnose the problem. An increased blood supply to the cervix results in a hyperaemic appearance and more pronounced vascular patterns even in the presence of low-grade disease. Hormonal stimulation of the endocervical glands increases the production of mucous, which can obscure the view of the colposcopist. Decidualisation of the cervix is not uncommon and confers an appearance alarmingly similar to invasive cancer made even more confusing by the frequently associated abnormal cervical bleeding. This was a feature in 6.2% of pregnant women who underwent routine colposcopy in pregnancy.
The natural history of CIN in pregnancy
Reliable information on the natural history of cervical intra-epithelial abnormalities in pregnancy is difficult to obtain. Previous studies must be interpreted with a degree of caution because of methodological problems, which include inconsistent criteria for entry into the studies, variable intervals of follow-up, variations in rates of colposcopically directed biopsies and the consequent effect of any biopsies on the natural history. Attempts to examine this question usually have involved the retrospective review of relatively small numbers of women and estimation of any progression and regression must again be done cautiously ( Fig. 1 ). In addition, considerable degree of subjectivity exists in the histological and diagnosis of CIN; diagnostic inconsistencies have been reported. These are more frequently associated with minor abnormalities than CIN III, which is considered a more robust diagnosis. Studies of the variability of cytological diagnosis have shown similar inconsistencies in the grading of cervical smears. Despite these limitations, one clear and consistent finding is evident; the development of invasive cancer during pregnancy is rare, and conservative management is safe provided close follow-up is carried out by experienced colposcopists.
The natural history of CIN in pregnancy
Reliable information on the natural history of cervical intra-epithelial abnormalities in pregnancy is difficult to obtain. Previous studies must be interpreted with a degree of caution because of methodological problems, which include inconsistent criteria for entry into the studies, variable intervals of follow-up, variations in rates of colposcopically directed biopsies and the consequent effect of any biopsies on the natural history. Attempts to examine this question usually have involved the retrospective review of relatively small numbers of women and estimation of any progression and regression must again be done cautiously ( Fig. 1 ). In addition, considerable degree of subjectivity exists in the histological and diagnosis of CIN; diagnostic inconsistencies have been reported. These are more frequently associated with minor abnormalities than CIN III, which is considered a more robust diagnosis. Studies of the variability of cytological diagnosis have shown similar inconsistencies in the grading of cervical smears. Despite these limitations, one clear and consistent finding is evident; the development of invasive cancer during pregnancy is rare, and conservative management is safe provided close follow-up is carried out by experienced colposcopists.
Women with high-grade abnormalities
In a 5-year period at one centre, 34 women were evaluated during pregnancy, of which 26 also had postpartum evaluation. One woman was treated (cone biopsy) during pregnancy for disease suspicious for microinvasion. She was disease-free postpartum. Of the remaining 25 women, 20 (80%) had persistent disease, two (8%) had progressive disease postpartum and three (12%) resolved without treatment.
In another study, follow-up was described in 153 women who underwent colposcopically directed biopsy in the ante-partum period ; 82 with CIN II and 71 with CIN III. The regression rates were 68% and 70% among CIN II and CIN III patients, respectively. Seven percent of patients with CIN II progressed to CIN III on postpartum evaluation. Twenty-five percent of those patients with CIN II and 30% of those with CIN III remained the same postpartum.
A total of 208 pregnant women with an abnormal cytology were assessed over a 10-year period at another centre, 78 of whom were histologically proven to have CIN. All patients were followed up every 8–10 weeks by cytology and colposcopy during pregnancy and reassessed 8–12 weeks postpartum. Persistent high-grade CIN was diagnosed in the postpartum period in 30 cases (38.4%), with regression to CIN 1 in the remaining 48 cases. No case of invasive disease developed during the follow-up period in these pregnant patients.
A total of 78 women with histologically proven CIN in pregnancy were reviewed by colposcopy at 8–12 week intervals in a 5-year review : 36 (46.2%) with high grade CIN (CIN 2–3) and 42 (53.8%) with CIN 1. In women with CIN 2–3, no invasion was suspected during pregnancy. At postpartum evaluation, there was no invasive or microinvasive cancer; 19 (52.7%) had persistent high-grade CIN and regression to CIN1 or less occurred in 17 (47.3%) cases. In the group of women with CIN 1, six (14.3%) progressed to CIN 2–3, seven (16.6%) had persistent CIN 1 and no residual CIN was diagnosed in 29 (69%) cases.
Women with low-grade abnormalities
Seventy-six women with squamous atypia encountered during routine Papanicolaou smear screening in pregnancy were seen during a 4-year period in one centre. All were evaluated with repeat cytology and colposcopy during pregnancy and again postpartum. A normal TZ was detected without evidence of intraepithelial neoplasia in 46 women. An atypical TZ was identified in 30 women; five cases were compatible with CIN 2 or CIN 3. Colposcopically directed biopsies were performed in 31 women, 29 of which were in the postpartum period and CIN was confirmed in 16 women. The authors concluded that colposcopy could be avoided during pregnancy for these women with repeat cytology alone being performed during pregnancy and careful revaluation carried out in the postpartum period.
A later study involved an analysis of the significance of atypical squamous cells of undetermined significance (ASCUS) Pap smears in pregnant and non-pregnant women over a 4-year period. A total of 198 women were diagnosed with ASCUS and 122 had available follow-up information. Of those included in the study, 58 were pregnant at the time of diagnosis and 64 were non-pregnant. Both groups were comparable in age but the pregnant women had biopsies deferred until the postpartum period. Thirty-two (50%) non-pregnant patients had CIN on histology compared with 15 of the 45 (33.3%) ( p = 0.019) pregnant patients. CIN 1 was detected in 12 of the 45 (26.7%) biopsies of the postpartum patients and 26 of the 64 (40.6%) ( p = 16) non-pregnant patients. High-grade CIN was found in 2 of the 45 (4.4%) postpartum patients versus 6 of the 64 (9.4%) ( p = 12) non-pregnant patients. One pregnant woman with ASCUS was found to have cervical cancer (2.2%). The authors concluded that the risk of cervical cancer following an ASCUS Pap smear was 2.2% in their pregnant population and concluded that ASCUS diagnosed on the Pap smear in pregnant women warranted colposcopy and close follow-up.
Once the initial colposcopy suggests low-grade CIN, however, the vast majority of women (85%) will either persist or regress spontaneously. In the group of 42 women with biopsy-proven CIN 1 in pregnancy, six (14.3%) progressed to CIN 2–3, seven (16.6%) had persistent CIN 1 and no residual CIN was diagnosed in 29 (69%) cases.
Colposcopic management of women with cervical cytology abnormalities during pregnancy
The indications for colposcopy for women with cytological abnormalities during pregnancy are the same as for non-pregnant women ( Fig. 2 ).
The timing of the colposcopic examination is usually early in the second trimester. This is usually a result of policies of routine smear testing at the initial antenatal booking visit and the interval to receipt of the result. For other referrals, colposcopy is often avoided by convention during the first trimester; however, this practice is without evidence. Colposcopy services, however, must balance the risk of delayed diagnosis of an invasive cancer against the increased worry of the examination precipitating bleeding in early pregnancy and the resultant diagnostic confusion of threatened miscarriage and possible need for anti-D prophylaxis.
Colposcopy plays an important role in the evaluation of women with suspected cervical abnormalities. It allows the identification of the site of the abnormality as well as an estimation of the grade of abnormality including the presence or absence of features suggestive of invasive cancer. For pregnant women, the primary objective of the colposcopic examination is to rule out a diagnosis of invasive cancer. Colposcopy without biopsy is more commonly applied to pregnant women than their non-pregnant counterparts. The diagnostic limitations of colposcopy have been well documented with a lack of correlation between the colposcopic and histological diagnosis. Quality assurance and training are therefore essential in maximising the opportunity for accurate diagnosis for these women.
As for non-pregnant women, dilute acetic acid is applied to the cervix under direct vision and colposcopists use a process of pattern recognition to discriminate between normal, low-grade and high-grade disease. This solution interacts with the cells’ protein structure and abnormal tissue shows up as a white area known as an atypical TZ. The intensity of the colour change as well as the sharpness of the margins help discriminate between grades of abnormality as does the rapidity with which the changes evolve and disappear.
A process of new blood vessel formation often accompanies high-grade CIN on the cervix. These give rise to patterns which are distinguishable as punctuation (end on blood vessels which look like a series of red dots) and mosaicism (a cobblestone or crazy paving type of appearance). In the presence of invasive disease, this process becomes more chaotic with the appearance of abnormal blood vessels. These have a curly or spiral shape in contrast to the tree-like root and branch-type normal vasculature, which occurs in response to infection. The appearance of these vessels can be enhanced by the application of a green filter, which excludes red light. The increased blood supply to the uterus in pregnancy results in increased vascular changes on the cervix, which can be a source of confusion for less experienced colposcopists. Finally, the location of the TZ can be checked using the application of Lugol’s iodine. This stains mature squamous epithelium a dark brown colour while the immature epithelium stains an orange colour.
Decidualisation of the cervix is a condition which can pose diagnostic problems for colposcopists. This is a benign change in the cervix and can be present in up to one-fifth of normal pregnant women. This condition commonly presents with bleeding during pregnancy and the difficulty is that it can have an appearance which is alarmingly like that of invasive cancer. It may look like cancer but is not cancer. In most cases, no treatment is required other than light cauterisation to control bleeding. Any suspicious lesion should be biopsied to rule out malignancy. Once the diagnosis is confirmed, the pregnancy should proceed as normal with complete resolution of the condition in the postnatal period.
Clear documentation is vital and must include a description of the atypical features identified, the location of both the TZ and the squamocolumner junction (SCJ) as well as the colposcopic impression. Cervical imaging using either digital photography or videocolposcopy has been shown to represent what is seen at colposcopy and is useful for patient care as well as for teaching and audit. This is of particular relevance for the colposcopic surveillance of pregnant women.
One difference between colposcopy practice in pregnant and non-pregnant women is the relatively infrequent practice of confirmatory biopsy with rates as low as 6%. The aim of directed biopsy is to sample an area, which is indicative of the most abnormal area to provide confirmation of the colposcopic impression ; the accuracy of these biopsies is dependent on the target site chosen by the colposcopist. This avoidance of biopsy does not have a basis in evidence. In a series of 149 women, little morbidity was in fact documented for colposcopically directed biopsies in pregnancy. No massive haemorrhage, premature labour, abortion or infection occurred. There was a high level of concordance between the colposcopic impression and the tissue biopsies (79%). By contrast, the practice of more selective biopsy can result in underestimating the severity of disease; 35% in one series of 34 women, with invasive cancer diagnosed during the postpartum period in two women. Cytological follow-up is equally unreliable with 37.5% (18/48) of women in another study with normal or low-grade repeat smears during pregnancy having CIN 3 diagnosed postpartum. While many colposcopists in Britain reserve diagnostic punch biopsy for women with suspected high-grade or invasive disease, routine biopsy of colposcopically atypical areas has been suggested as the standard of care by others.
The potential utility of HPV triage in women with ASCUS diagnosed in pregnancy has been studied. The prevalence of HPV DNA in pregnant women (88.6%) was similar to that found in ASC in non-pregnant women of the same reproductive age-group (83.8%), and the high-risk types accounted for the vast majority of cases (83.9%). These findings suggest that HPV testing may have limited utility in effective management of these patients.
Irrespective of whether an initial biopsy is performed, close supervision with expert colposcopy is required during pregnancy and in the postnatal period. The frequency of visits depends on the suspected grade of disease. Following initial assessment, a schedule should include review in the early third trimester and 8 weeks postnatal for women with satisfactory colposcopy and suspected low-grade disease. For women with suspected high-grade disease, an additional assessment at 32 weeks is required. Close communication with the attending obstetrician is recommended and may reduce the risk of default ( Fig. 3 ).
The availability of colposcopy and the ability to be able to distinguish between women with and without cancer can avoid conisation in the vast majority of pregnant women. Conisation is still sometimes required and is indicated for women with suspected high-grade disease and an unsatisfactory colposcopy as well as for women with suspected invasive cancer. These biopsies cannot be considered as curative during pregnancy because of the high percentage of positive margins and the relatively high incidence of residual disease postpartum. In a series of 55 knife-cone biopsies in pregnancy, 44 women had high-grade CIN and three had invasive cancer, giving incidence figures in the pregnant populations of 0.22% for CIN and 0.015% for cancer. The associated pregnancy loss in the second trimester was 4.2% with a preterm labour rate of 12.5%. The vaginal delivery rate was 52% and the LSCS rate was 48%. Ablative treatment during pregnancy is not recommended.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
