Management of cervical intraepithelial neoplasia (CIN) needs to protect women at risk from developing cervical cancer and to avoid over-treatment as well as obstetrical complications in women undergoing invasive treatment. Strong evidence shows that CIN3 is a true precursor and must be treated, whereas CIN1 lesions do not benefit from immediate surgery and should be followed conservatively. Although the clinical course of CIN2 differs from CIN3, it should be treated the same way for legal reasons. Colposcopy plays a central role in selection of patients and treatments. Treatment of CIN2 and 3 should be excisional. Large loop excision of the transformation zone, high-frequency-needle or laser conisation are equally good, whereas cold-knife conisation is associated with an excess risk for subsequent obstetrical complications. Human papillomavirus testing and cytology at 6 months seems to be the best post-treatment monitoring, although this needs to be confirmed by randomised-controlled trials. Future research needs to focus more on how the quality of colposcopy and the overall management concept determines the clinical outcome instead of exploring the role of single technical methods. Furthermore, it seems to be necessary to evaluate the best management of CIN2 in young and in vaccinated women.
Introduction
The major ethical and medical aim of all management options for cervical intraepithelial neoplasia (CIN) is the prevention of cervical cancer. The second most important aim is to achieve this with as little harm to the woman as possible. Finally, management of CIN should be cost-efficient without compromising any of its two major aims. As only a minority of CIN will progress to cervical cancer, and because any kind of invasive treatment is associated with adverse events, good management of CIN is based on a good selection of women who are really in danger of developing cervical cancer and need to be treated, compared with those who are not at risk of cancer and should be protected from over-treatment.
With the deep understanding of the human papillomavirus (HPV)-dependant genesis of cervical cancer, we are able to calculate risks of progression of CIN quite accurately, and have learned about minimum periods for each step of cervical carcinogenesis. One important principle, however, in the management of CIN is over-treatment. We know that, even with the best available selection, we will treat CIN lesions that would never progress to cancer, and this is acceptable as long as it is not excessive over-treatment. State-of-the-art treatment of CIN has moderate side-effects, whereas cervical cancer is a deadly malignant disease. Selection of women in need of protection from cervical cancer depends mainly on the grade of CIN, the type of the transformation zone, and age of the woman.
It is well-established that persistent infection of certain HPV is an essential precondition in the genesis of cervical cancer. Virtually all primary cervical cancers contain HPV DNA. HPV-induced carcinogenesis, however, is a rather slow process, with at least four important steps: incident infection with a high-risk HPV-type; establishment of a persistent HPV infection; transformation of human keratinocytes in pre-invasive neoplastic cells; and malignant transformation of intraepithelial neoplasia into invasive cancer.
The process is reversible at any pre-invasive stage. Most genital HPV infections are transient, and more than 90% will be cleared spontaneously within 18 months in young women. Persistent HPV infections and low-grade intraepithelial neoplasia will regress without any treatment within 2–5 years in 60–80% of all cases. Spontaneous resolution is observed in at least 40% of lesions with biopsy-proven moderate dysplasia (cervical intraepithelial neoplasia grade II). Only a 20% minority of CIN2 will progress to CIN3, and less than 5% to invasive cancer.
It is quite obvious that the morphology-based grading of pre-invasive cervical neoplasia does not always reflect the risk of progression to invasive disease. We have learned from observational studies and from large randomised-controlled HPV vaccine trials that, in young, HPV-naive women at baseline, incident HPV infections may be associated with CIN1 and even CIN2 within a few months. Although these lesions cannot be distinguished by cyto- or histomorphology from persisting CIN1 or CIN2, the clinical course seems to be different, with a much higher likelihood of regression. Although the natural history of CIN2 resembles that of low-grade lesions in most cases, it is regarded as a high-grade lesion in most clinical management guidelines for legal reasons. There is sufficient evidence that CIN3 is a true precursor of cervical cancer.
At the National Women’s Hospital, Auckland, New Zealand, treatment of CIN3 was withheld from a substantial number of women between 1965 and 1974 as part of an unethical clinical study. During long-term follow up, 31.3% of women managed with punch biopsies only developed invasive cancer. In the subset of 92 such women who had persistent disease within 24 months after biopsy-proven CIN3, invasive disease occurred even in 50.3% (37.3–64.9).
Principles, outcomes and standards of observational management
Excisional treatment of CIN is associated with a significantly increased risk of preterm delivery, late miscarriage, and perinatal mortality in subsequent pregnancies. Therefore, it should be used only in well-selected women with a clear indication for invasive treatment.
As pointed out, CIN1 is common, especially in adolescents and young women. In many cases, these lesions are just cellular changes of self-limiting, permissive HPV infections with a high rate of spontaneous regression. Even in persisting CIN1, the risk of progression is low, and most of these lesions will finally undergo spontaneous regression. The European guidelines and the 2006 American Society for Colposcopy and Cervical Pathology (ASCCP) consensus guidelines recommend observational management in women with CIN1. Compared with the 2001 ASCCP guidelines, cytological follow up became the only recommended management option in the 2006 recommendations, regardless of whether the colposcopic examination is satisfactory, for women with CIN1 who have a low-grade referral cervical cytology. Treatment of CIN1 is particularly discouraged in adolescents.
For the management of CIN, it is important that cytologic low-grade squamous intraepithelial lesion (LSIL) is not equivalent to histologic CIN1 and cytologic high-grade squamous intraepithelial lesion (HSIL) is not equivalent to histologic CIN2 and 3. Because the prevalence of CIN2 or greater was 12–16% at initial colposcopy among women with LSIL in randomised-controlled trials (RCTs), the ASCCP guidelines recommend colposcopy for managing all women with LSIL, with the exception of adolescents who should be followed with annual Pap smears. This recommendation, however, is questioned by the results of a large RCT in the UK. The Trial of Management of Borderline and Other Low Grade Abnormal Smears (TOMBOLA) compared cytological surveillance with immediate referral to colposcopy in the management of women with low-grade cervical abnormality in 4439 eligible women. Immediate referral to colposcopy detected more CIN2 or greater lesions at baseline, but the trial found no significant difference in the cumulative incidence during 3 years’ follow up. Immediate colposcopy, however, resulted in a large number of referrals of women with no relevant cervical disease.
Although HPV testing proved to be efficient in the triage of ASCUS/borderline cytology, HPV testing is of little use in LSIL cytology because about 80% will be associated with HPV.
From the view of a colposcopy clinic, women transferred because of borderline or LSIL cytology should undergo colposcopic assessment with biopsies. The TOMBOLA trials could show, with a high level of evidence, that immediate large loop excision of the transformation zone (LLETZ) will not increase the rate of CIN2 or CIN3 compared with targeted punch biopsies by experienced colposcopist, but results in over-treatment and unnecessary morbidity in healthy women. Another RCT in Canada and Brazil compared immediate LLETZ and expectant management in women with biopsy-proven CIN1. No significant difference was found in the risk of progression to CIN2 or 3 between the two study arms. The authors concluded that expectant management of biopsy-proven CIN1 up to 18 months is a reasonable management strategy. Both RCTs support the broad consensus that, in women with biopsy-proven CIN1 and satisfactory colposcopy, invasive treatment is not indicated. These women should be followed with intensified screening. Because the risk of underlying high-grade disease is low, even in women with LSIL and unsatisfactory colposcopy, most guidelines also recommend observational management in these cases. Graphic 1 summarizes the management options of LSIL ( Graphic 1 near). However, in women referred to colposcopy because of HSIL cytology, who receive a histologic diagnosis of CIN1 in corresponding targeted biopsies, most guidelines recommend a diagnostic excision of the transformation zone because the risk of underlying high-grade disease is high.
In women younger than 25 years, the incidence of invasive cervical cancer is low, whereas the rates of HPV infections, CIN1 and CIN2 are high. As most of these lesions will regress spontaneously, many national screening programmes recommend screening only in women who are 25 years or older to protect younger women from over-treatment. In countries with an earlier start of screening, this should be kept in mind. The 2006 ASCCP guidelines encouraged expectant management of CIN2/3 in adolescents. In our own experience, expectant management of biopsy-confirmed CIN2 in women with satisfactory colposcopy is safe and avoids over-treatment in women younger than 30 years. HPV-genotyping seems to be useful in predicting the clinical course of CIN2. Although CIN2 lesions associated with HPV16 carry a high risk of progression, most CIN2 associated with other HPV types will regress spontaneously. Expectant management of CIN2 outside of trials, however, should be restricted to women younger than 25 years with a good compliance, and the follow-up should be controlled by an experienced colposcopy clinic.
During pregnancy, invasive treatment of any grade of CIN should be postponed until 2 months after delivery. Invasive disease must be excluded by colposcopy. If biopsies need to be taken at the 18 th week of gestation or later in pregnancy, loop biopsies are more appropriate than standard punch biopsies to obtain a histological specimen with sufficient stroma. When invasive cancer can be excluded with high reliability, the remaining risk for CIN3 lesions to progress to malignancy during pregnancy is small, whereas the fetal and maternal risks related to invasive treatment of the cervix are high, and complete excision of lesions cannot be achieved in almost half of the cases.
Principles, outcomes and standards of observational management
Excisional treatment of CIN is associated with a significantly increased risk of preterm delivery, late miscarriage, and perinatal mortality in subsequent pregnancies. Therefore, it should be used only in well-selected women with a clear indication for invasive treatment.
As pointed out, CIN1 is common, especially in adolescents and young women. In many cases, these lesions are just cellular changes of self-limiting, permissive HPV infections with a high rate of spontaneous regression. Even in persisting CIN1, the risk of progression is low, and most of these lesions will finally undergo spontaneous regression. The European guidelines and the 2006 American Society for Colposcopy and Cervical Pathology (ASCCP) consensus guidelines recommend observational management in women with CIN1. Compared with the 2001 ASCCP guidelines, cytological follow up became the only recommended management option in the 2006 recommendations, regardless of whether the colposcopic examination is satisfactory, for women with CIN1 who have a low-grade referral cervical cytology. Treatment of CIN1 is particularly discouraged in adolescents.
For the management of CIN, it is important that cytologic low-grade squamous intraepithelial lesion (LSIL) is not equivalent to histologic CIN1 and cytologic high-grade squamous intraepithelial lesion (HSIL) is not equivalent to histologic CIN2 and 3. Because the prevalence of CIN2 or greater was 12–16% at initial colposcopy among women with LSIL in randomised-controlled trials (RCTs), the ASCCP guidelines recommend colposcopy for managing all women with LSIL, with the exception of adolescents who should be followed with annual Pap smears. This recommendation, however, is questioned by the results of a large RCT in the UK. The Trial of Management of Borderline and Other Low Grade Abnormal Smears (TOMBOLA) compared cytological surveillance with immediate referral to colposcopy in the management of women with low-grade cervical abnormality in 4439 eligible women. Immediate referral to colposcopy detected more CIN2 or greater lesions at baseline, but the trial found no significant difference in the cumulative incidence during 3 years’ follow up. Immediate colposcopy, however, resulted in a large number of referrals of women with no relevant cervical disease.
Although HPV testing proved to be efficient in the triage of ASCUS/borderline cytology, HPV testing is of little use in LSIL cytology because about 80% will be associated with HPV.
From the view of a colposcopy clinic, women transferred because of borderline or LSIL cytology should undergo colposcopic assessment with biopsies. The TOMBOLA trials could show, with a high level of evidence, that immediate large loop excision of the transformation zone (LLETZ) will not increase the rate of CIN2 or CIN3 compared with targeted punch biopsies by experienced colposcopist, but results in over-treatment and unnecessary morbidity in healthy women. Another RCT in Canada and Brazil compared immediate LLETZ and expectant management in women with biopsy-proven CIN1. No significant difference was found in the risk of progression to CIN2 or 3 between the two study arms. The authors concluded that expectant management of biopsy-proven CIN1 up to 18 months is a reasonable management strategy. Both RCTs support the broad consensus that, in women with biopsy-proven CIN1 and satisfactory colposcopy, invasive treatment is not indicated. These women should be followed with intensified screening. Because the risk of underlying high-grade disease is low, even in women with LSIL and unsatisfactory colposcopy, most guidelines also recommend observational management in these cases. Graphic 1 summarizes the management options of LSIL ( Graphic 1 near). However, in women referred to colposcopy because of HSIL cytology, who receive a histologic diagnosis of CIN1 in corresponding targeted biopsies, most guidelines recommend a diagnostic excision of the transformation zone because the risk of underlying high-grade disease is high.
In women younger than 25 years, the incidence of invasive cervical cancer is low, whereas the rates of HPV infections, CIN1 and CIN2 are high. As most of these lesions will regress spontaneously, many national screening programmes recommend screening only in women who are 25 years or older to protect younger women from over-treatment. In countries with an earlier start of screening, this should be kept in mind. The 2006 ASCCP guidelines encouraged expectant management of CIN2/3 in adolescents. In our own experience, expectant management of biopsy-confirmed CIN2 in women with satisfactory colposcopy is safe and avoids over-treatment in women younger than 30 years. HPV-genotyping seems to be useful in predicting the clinical course of CIN2. Although CIN2 lesions associated with HPV16 carry a high risk of progression, most CIN2 associated with other HPV types will regress spontaneously. Expectant management of CIN2 outside of trials, however, should be restricted to women younger than 25 years with a good compliance, and the follow-up should be controlled by an experienced colposcopy clinic.
During pregnancy, invasive treatment of any grade of CIN should be postponed until 2 months after delivery. Invasive disease must be excluded by colposcopy. If biopsies need to be taken at the 18 th week of gestation or later in pregnancy, loop biopsies are more appropriate than standard punch biopsies to obtain a histological specimen with sufficient stroma. When invasive cancer can be excluded with high reliability, the remaining risk for CIN3 lesions to progress to malignancy during pregnancy is small, whereas the fetal and maternal risks related to invasive treatment of the cervix are high, and complete excision of lesions cannot be achieved in almost half of the cases.
Principles, standards and outcomes of surgical treatment modalities
Surgical treatment of CIN3 is the accepted treatment standard worldwide. Conisation and LLETZ are the most common surgical methods. In many countries, invasive treatment is obligatory for CIN3 and CIN2, although the risk of progression to invasive disease is low in CIN2, and many CIN2 lesions will regress spontaneously in younger women.
The primary goal in the management of CIN2 and 3 is to prevent the development of invasive cancer by complete surgical destruction of all neoplastic tissue. To achieve this with as little loss of cervical function as possible is an important secondary goal.
The healthy uterine cervix protects the uterine and abdominal cavities from ascending infections and plays an important role in the regulation of fertility and during pregnancy. As the mean age of women with CIN3 is about 31–34 years in most populations, destructive treatment of the cervix may affect fertility and increase significantly the number of late miscarriages, premature labour, preterm delivery, and perinatal mortality in subsequent pregnancies. This high risk of obstetrical complications after invasive treatment of the uterine cervix was confirmed by two meta-analyses.
Colposcopy is the single method of identifying the exact topography of cervical lesions, and thus plays a crucial role in selecting the appropriate surgical treatment of individual CIN3. The target host cells of primary HPV infection are the basal cells that are exposed superficially at the squamocolumnar junction. Furthermore, it seems to be one peculiarity of the cervical transformation zone that its metaplasia cells are especially prone to oncogenic transformation induced by HPV because these cells need to provide different pathways of cell differentiation. This explains the well-known observation that almost all high-grade cervical lesions are located within the cervical transformation zone, whereas vaginal intraepithelial neoplasia is rare, although vagina and uterine cervix are exposed to HPV at a similar rate.
To achieve high cure rates in the treatment of premalignant disease of the cervix, it is necessary to destroy the complete transformation zone and any neighbouring lesions. As the extension of individual CIN3 lesions and transformation zones differ significantly, colposcopically guided surgical treatment will remove neoplastic tissue with high accuracy and avoid unnecessary destruction of neighbouring healthy tissue. It seems to be a matter of course for Anglo-Saxon countries that colposcopy plays a central role in the management of atypical screening results and guidance of invasive treatment, but this is not reality in many other countries. A retrospective nationwide study in Germany found that more than 90% of women who were treated because of LSIL or HSIL smears never had colposcopy assessment and underwent directly diagnostic cold-knife biopsy or even hysterectomy. This management resulted in significant over-treatment and increased costs. This is in line with the TOMBOLA evaluation that colposcopic evaluation of LSIL smears are cost efficient.
Any method of treating CIN 3 carries a small risk of subsequent development of invasive cancer. This event seems to occur less frequently after excisional treatment with cold-knife-, needle-, laser-cone biopsy or LLETZ than after destructive treatment using laser vaporisation, electrocoagulation diathermy, cryotherapy or cold coagulation. Therefore, and because of the better histological evaluation, excisional methods are the standard treatment for CIN. The use of destructive techniques should only be applied for extended and multifocal lesions.
No single excisional method could ever prove superior cure rates in randomised, multicentre trials, and the multiple smaller observational or single-centre studies differ significantly in study design and do not allow for a general recommendation of one method of surgical treatment of CIN3. It is possible, however, to recommend the best surgical method for each individual CIN3 based on the colposcopical categorisation of the transformation zone and the evidence-based risk–benefit profile of available surgical methods.
Surgical treatment of CIN requires the complete removal of the transformation zone. The International Federation for Colposcopy and Cervical Pathology defined three different types of TZ:
Type 1 The complete TZ is located on the ectocervix
Type 2 The TZ extends into the endocervical canal but the upper border/Squamous columnar junction is visible on colposcopy
Type 3 The TZ extends into the endocervix and the upper border is not visible on colposcopy
In all types of transformation zone, the ectocervical or vaginal border of the lesion, and in types 1 and 2, even the endocervical extension can be identified by colposcopy. To achieve a full thickness resection of a type 1 transformation zone with clear basal margins, the cutting depth for the external orifice must be 8 mm, whereas 5 mm may be sufficient for the periphery of the transformation zone.
LLETZ is the best suited treatment for CIN located in transformation zone type 1 and 2. The loops are tailored for an optimal bowl-like resection of the transformation zone. The size of the transformation identified on colposcopy determines the size of the loop. The cutting effect is achieved by light arches that spark between the 0.2-mm thin wire and the tissue. Modern high-frequency (HF-) generators regulate electronically the needed optimal cutting currency. Properly performed LLETZ causes only minimal thermonecrosis that marks the definite surgical resection margin. The obtained samples allow a full histopathological review. The outpatient procedure is safe, easy, can be carried out under local anaesthesia, and causes little bleeding and complications. Haemostasis can be achieved easily by application of Monsell’s solution. Transformation zone types 1 and 2, with minimal extension to the endocervix, can be resected completely with a loop in one piece, and the treated cervix will heal without significant scarring or deformations in most cases (see Picture 1 near). Transformation zone type 2 that extends 7–10 mm into the endocervix may require a second resection of the adjacent canal with a smaller loop to achieve clear endocervical margins.
Full access? Get Clinical Tree
