Regression rate of high-grade cervical intraepithelial lesions in women younger than 25 years

Since introduction of cervical cancer screening in several, especially high-income, countries, the incidence and mortality of cervical cancer decreased markedly.

In Germany, cervical cancer screening was introduced in 1971, offering women yearly examinations of the cervix by cytology, or so-called “pap smear.” Within the last 15 years, the incidence and mortality of cervical cancer is largely stable [1]. The average age at initial diagnosis of invasive cervical cancer is 55 years [1].

The primary goal of secondary prevention of cervical cancer is to discover precancerous lesions of the cervix in order to prevent progression to invasive cancer.

The main cause of cervical cancer and its precursors is a persistent infection with human papilloma viruses (HPV) [2]. An asymptomatic infection is very common in young women, especially younger than 25 years, and in most cases, the HPV infection is recognized by the immune system and clears spontaneously without consequences [3]. The lifetime probability of HPV infection in sexual active adults is reported to be about 85–90% [3]. A persistent infection with high-risk HP viruses, especially the high-risk types 16 or 18, has a substantial risk to progress to cervical intraepithelial lesions within months to years, and to progress to cervical cancer within 7–15 years [4].

The course of HPV infection in squamous epithelium is well examined:

After sexual intercourse, women may acquire an HPV infection. Due to micro-lesions of the epithelium, the virus particles can “reach” the basal layer of the squamous epithelium and infect these basal cells [5]. This is called latent infection and can last for many years.

During permissive infection, the amplification cycle of the virus particle, including infectious descendants, takes place in epithelium cells [5]. These cells are called koilocytes. The viral reproduction cycle and the differentiation of the epithelium are aligned to each other.

In most cases, the human immune system is able to clear the HPV infection within 1–2 years [3].

About 10% of infected individuals do not clear the virus and develop persistent infection: HPV persists in the squamous cells and is replicated geared to the cell cycle of the epithelium [3].

In some cases, it could lead to a transforming infection: a turnover of normal to dysplastic or malignant cells. Risk factors are compromised immune system, HIV infection, smoking cigarettes and immunosuppressive medications [6].

The precancerous lesions of squamous epithelium, or so-called cervical intraepithelial lesions (CIN), can range from mild (CIN 1), moderate (CIN 2) to severe (CIN 3) lesions and carcinoma in situ. Adenocarcinoma in situ (AIS) is the glandular precursor for invasive endocervical adenocarcinoma.

International guidelines recommend a surgical treatment for high-grade intraepithelial lesions (CIN 2/3) and a watch-and-see-strategy for low-grade lesions (CIN1).

German guidelines suggest waiting up to 24 months in cases of CIN 1 by performing colposcopy every 6 months. CIN 2 can be monitored for at least 12 months as well. In case of persistence or progression to CIN 3, treatment by surgery is recommended [7].

The recommended treatment of histologically proven CIN 3 is surgery, only in young individuals a short-term observational period is possible [7].

The standard procedure is conization of the cervix uteri. Nowadays, a loop electrosurgical conization under colposcopic guidance is the gold standard (LEEP = loop electrosurgical excision procedure). Alternative procedure is laser conization. CIN 1–2 may also be treated by laser coagulation of the cervical lesion after biopsy for histology [7]. Conizations performed by cold knife are obsolete.

Specific complications are bleeding, cervical stenosis or shortened cervix during following pregnancies with the risk of premature rupture of membranes and preterm deliveries [8‐10].

Especially in young women who have not decided their family planning, yet, surgical treatment should be indicated with restriction and caution.

As regression rates in mainly CIN 2 lesions are high, it is important to find special predictors, which can accurately predict which lesions may regress or not.

Table 1 summarizes some studies about observational management of CIN 2 in young women. The observed regression rates were between 39 and 71.1%, whereas the progression rates were about 8.3–16.6% [11‐14].

The discrepancy in results may be explained by several factors:

In most studies focused on the regression rates, there was no differentiation between the CIN 2 and the CIN 3 group. Furthermore, p 16 immunohistochemistry was not performed. This can lead to misleading results in histology, as inflammation was misinterpreted as dysplasia in some cases.

In many studies, there were only a restricted number of patients, which makes it even more difficult to determine the numbers of regressions in an appropriate way.

Taken all together, regression rates and their appropriate predictors have to been analyzed in an accurate way to find out whether a conservative approach with short-term controls may be safe in young women and could help to avoid unnecessary surgery to preserve the cervix and reduce pregnancy complications.

This is a retrospective cohort study.

Nowadays, most of the women younger 25 years have not decided their family planning, yet. Therefore, uterine surgery such as loop conizations should be recommended restrictively and cautiously. Thus, “watch and wait” strategies for cervical dysplasia especially in young women need to be discussed.

So far, there are only a few data available about regression rates, progression rates and times of observation periods for CIN 3 in the young age group, in contrast to several studies concerning regression rates in CIN 2.

The regression rates in CIN 2 lesions in previous studies vary from 39 to 71% (see Table 1), whereas in CIN 3 lesions, the rates vary from 1.3 to 38% [11‐18].

A large Australian retrospective cohort study published by Munro et al. 2016 analyzed conservative management of 924 Western Australian women aged 18–24 years diagnosed with CIN2 on cervical biopsy [12]. In their cohort, the 2-year regression rate for CIN2 was 59.5% and the progression rate was 16.4% (CIN 3 and AIS). None of the women progressed to invasive cancer.

In the same year, Loopik et al. published another retrospective cohort study (n = 211) of women younger than 25 years with CIN2 managed conservatively [13]. The regression rate was even better with 71.1%) showed regression, and 16.6% of the young women progressed (no cancer), with a median follow-up of 15.1 months. Smoking was a risk factor for progression (hazard ratio 2.40, p = 0.006).

A prospective study by Moscicki et al. examined the progression and regression including risk factors of young women (13–24 years) and CIN 2, n = 95 [14]. They found a 2-year regression rate of 63% and 68% after 3-year follow-up. The progression to CIN 3 was 15% and similar to the above studies. Non-progression and regression were associated with non-persistence of HPV infection (hazard ratio 0.40; 95% CI 0.22–0.72) or oral contraceptive use (hazard ratio 0.85; 95% CI 0.75–0.97).

Another small cohort study evaluated regression rates among adolescents (aged < or = 21) with CIN2 and managed expectantly [11]. 36 young women could be followed conservatively and regression after a median follow-up time of 378 days was documented in 14 (39%). The authors defined regression as complete regression to normal cytology and biopsy/colposcopy. 19 patients had a CIN 1 or mildly abnormal cytology results and only 3 patients developed a persistence or progression. Therefore, in total, 92% of women younger than 21 and CIN 2 had regressive results during conservative management.

In our cohort study of 151 cases, the regression rate of CIN 2 was 88% (n = 30) and even 76% (n = 26) for complete regression, which is higher than the published data so far, and without any case of progression during the observation period of up to 5 years. Reasons for these might be a very high standard and experience of the examiners at our colposcopy unit and good education and reassurance to the patients that the expectative management is safe, and they have a good chance to clear their HPV infection/lesion.

For CIN 3, little conclusive data are available—the published regression rates vary between 1.3 and 38% [15‐18]. In our cohort, 29% of the patients younger than 25 years (n = 34) developed regressive results after an observation period of up 5 years, and even 14% (n = 16) had complete regression of cervical dysplasia. We encountered no case of progression and of invasive disease during observation time.

Motamedi et al. published 2015 a retrospective analysis of 635 cases with CIN 3. Their regression rate of CIN 3 was 1.3% (n = 8). But, they detected 12 invasive carcinomas after conization in the cone tissue. Therefore, the rate of undetected cervical cancer was higher than the regression rate and they concluded that observational management of CIN 3 is not justifiable [15]. It should be noted that in this study, the median women’s age was 32 years, and there is no age-related analysis concerning regression rate in the younger age group available.

Moreover, many studies did not distinguish between CIN 2 and CIN 3. Often, they did not use immunohistochemistry for histological diagnoses, which might cause mistakes in classification. We must remark, in our cohort also 19 cases could not be distinguished between CIN 2 or 3, so we treated them as CIN 3. One cannot rule out the possibility that the patients “just” had a CIN 2, so the rates might be biased.

Munk et al. documented a regression rate of CIN 2–3 from 5% after 9 weeks observation up to 38% for a longer period but had a small sample size of 61 patients in total [16].

The aim of the study was to evaluate the hypothesis that “curative” punch biopsies are the reason for CIN 3 regression. With a rising regression rate by longer observation periods and regression despite positive resection margins of the cervical biopsies, they could disprove this hypothesis.

A meta-analysis done by Zhang et al. found regression rates for CIN 2 of 50.85% and of 36.31% for a group of CIN 2/3 (without differentiation of those) [17]. The median age was 28.23 years. Therefore, the trend of these results matches with our regression rates and confirm our data quite nicely.

Nonetheless, our results also show that the chance of regression of CIN 2 is much higher than CIN 3 (p < 0.001). CIN2 are less aggressive and rarely end in cervical cancer, so it has a much higher chance to regress spontaneously [18]. In our hands, young women with CIN 2 never get the indication for surgery right away but always are treated observationally, to the point that in our hands, immediate resection of CIN 2 in young women without risk factors is considered unnecessary surgical treatment. Based on the results of this study in our institution, also CIN 3 can be managed observationally in young women without any risk factors and informed consent, because the rate of regression in a third of those women is quite high and justifies the “wait and control” management. The reliability of the women and short-term follow-up examinations are precondition for this kind of management.

Looking at the time interval for regression, the retrospective cohort study published by Munro et al. 2016, (n = 924) evaluated a regression rate of CIN 2 of 59.5% within 2 years. These results perfectly match with our findings: we exploited a regression of CIN 2 in 61.7% (n = 21) of our cases within 2 years [12].

Lee et al. 2018 also analyzed the outcomes of conservative management of CIN 2 and 3 in young women and found regression rate (partial and complete) of CIN 2 of 74.7% (n = 74), which is a just a bit lower than the regression rate of 88% in our cohort [19]. The median time to regression was 10.8 months. In this study, the regression rate for CIN 3 was 21.6% (n = 11), but the median time to regression was not reached. In our study, the spontaneous regression of CIN3 was quite similar with 19.6% within 2 years (n = 23) and 24.8% within 3 years (n = 29).

To summarize, for young patients with CIN3 without any risk factors, an observation period should be recommended with informed consent of the patient. But for this conservative management, mechanisms are necessary to ensure that these women return to their follow-up, at least every 3–6 months.

Due to small sample sizes of published data to this point, larger studies are still required.

In total, the regression rate (partial and complete) for CIN 2 was 88.2% (n = 30) and significantly higher than the regression rate (partial and complete) for CIN 3 with 29.1% (n = 34), (p = 0.0002).

Within 2 years of a “wait and see” approach, 70% (n = 21) of CIN 2 and 68% (n = 23) of CIN 3 showed either a partial or even complete regression. None of the patients in this cohort (n = 151) developed progression of cervical dysplasia or even progressed to cervical cancer.

This allows us to prolong the observation period and the follow-up interval in patients with initial diagnosis of CIN 2 younger than 25 years, in order to reduce the necessity of surgical treatment of the uterine cervix. We suggest a follow-up including colposcopy ± biopsy to be repeated every 6 months for 2 years.

Moreover, we should check the indication for immediate surgical treatment in women with initial diagnosis of CIN 3 without any risk factors and being younger than 25 years. The observational strategy is justified for at least one year under close supervision by an expert team with colposcopy + biopsy/HPV testing if needed every 3–4 months.

To predict the regression of CIN 2 and 3 in patients younger than 25 years while “wait and see”, we found three significant factors: the initial histological diagnosis (CIN 2 or 3), the clearance of HPV infection and prior HPV vaccination.

We should also add the repetition of HPV high-risk testing at the follow-up examinations to determine potential HPV clearance.

Larger prospective cohort studies or case–control studies to examine the spontaneous regression rates of severe cervical dysplasia are needed, including data about progression rates. Furthermore, increasing vaccination rates of HPV might influence the regression rates. These data are important for comprehensive patient information.