Manageable risk factors for progression of HPV-associated cervical intraepithelial neoplasia

Introduction

Cervical cancer remains one of the most significant medical and social problems of female reproductive health [1]. It is known that globally, one woman dies of this disease every 40 seconds [2]. Since 2008, the role of human papillomavirus (HPV) as the main etiological factor of cervical intraepithelial neoplasia (CIN) and invasive cancer has been proven. In this regard, the most promising strategy in cervical screening is considered to be a combination of HPV typing and cytological examination of cervical epithelial cells, regulated by the Russian clinical guidelines of 2020 and the Order of the Ministry of Health of the Russian Federation No. 1130n. According to these documents, in young patients with low-grade squamous intraepithelial lesions (LSIL) (signs of HPV infection, koilocytosis) or patients planning pregnancy, a wait-and-see approach is preferable with dynamic monitoring of the state of the cervix within 18–24 months in the form of cytological monitoring once every 6 months and HPV testing once a year. Surgical treatment is recommended in the absence of regression after 18–24 months^{1 2}.

A number of studies established that HPV persistence, CIN and cervical cancer development depend on the methylation of tumor growth suppressor genes. After introduction into the cell, HPV uses human methyl group donors (folate and vitamin B12), leading to a significant increase in the synthesis of E6 and E7 oncoproteins, which stimulate cell proliferation, prevent apoptosis of cancer cells, block the production of interferon, and disrupt the protective regulatory mechanisms of DNA repair, which in general destabilizes the genome. An optimal level of folate is required to reduce oncoprotein gene expression in HPV-infected cells [3][4].

It is known that actively proliferating cells (hematopoietic and epithelial cells) react sensitively to the disruption of replication that occurs with folate deficiency, which is accompanied by impaired hematopoiesis (anemia, thrombocytopenia, and leukopenia) and a high risk of malignancies, due to impaired proliferation of epithelial cells of the skin and mucosa. The role of low folic acid in increasing the risk of several other conditions, including disorders of embryogenesis, neurodegenerative and cardiovascular diseases, was also proven [5][6].

However, studies establishing associative links between the risk of developing or progressing CIN and serum folate concentrations are limited and presented only by foreign authors of the past decade [7-9].

Due to the high prevalence of HPV, the natural history of its course, the possibility of both regression and progression, and the search for informative prognostic markers, making it possible to form groups of "high" and "low" risk of HPV infection realization in cervical squamous intraepithelial lesions, are of scientific and practical interest.

The study aimed to identify informative markers of the progression of low-grade cervical squamous cell intraepithelial lesions.

Materials and Methods

A total of 127 patients were examined, who had received a positive HPV typing test for high carcinogenic risk (HCR) 6 months earlier during routine cervical screening in outpatient clinics in Rostov-on-Don. Forty-four women infected with HPV had no cervical intraepithelial malignant lesions of squamous epithelium (NILM), and 83 patients (65.4%) had LSIL. A wait-and-see approach was chosen for all patients in accordance with current clinical guidelines, namely a dynamic follow-up with follow-up consultations after 6 months of observation.

Analysis of HPV testing of 127 patients showed that the most common infection was with 16, 18, and 33 genotypes, which were detected in 90 women (70.9%) and accounted for 43.2%, 16.3%, and 17.1%, respectively, in the structure of HPV HCR.

To choose further management of 90 patients six months after the initial examination, a comparative analysis of the dynamics of laboratory and instrumental methods of cervical diagnosis (cervical smear cytology, molecular-biological HPV examination of cervical smear with genotyping and viral load, extended colposcopy, immunocytochemical testing of cervical scrape with detection of p16INK4a protein) was performed.

The results revealed significant differences in the outcomes of the dynamic follow-up, which determined the distribution of patients into three clinical groups.

Regression of the disease was detected in 43 out of the 90 patients (47.8%) in Group I. Group II consisted of 30 out of 90 patients (33.3%) with the persistence of the disease (absence of significant changes in the results of laboratory and instrumental methods of examination). Group III consisted of 17 out of 90 patients (18.9%) with the disease progression.

The authors analyzed the complaints in the study groups and the condition of the vaginal microbiota at the time of consultation, the age of the patients, sexual debut, the social status of the patients with the analysis of promiscuity and marital status, the use of contraceptive methods, the gynecological and obstetrical history, the history of extragenital diseases, and the serum level of vitamin B₉ concentration.

The study entry criteria included age 18 to 49 years old, a positive cervical HPV test for types 16, 18, and 33, the presence of cervical squamous intraepithelial lesion (LSIL) or its absence (NILM), and informed consent of the patient.

Exclusion criteria were verification of high-grade cervical intraepithelial lesion (HSIL), intravaginal administration of any drugs, administration of immunomodulators, antiviral and antibacterial drugs within 6 months, pregnancy and lactation, refusal to participate in the study.

For the liquid cytology of the cervix, cellular material was obtained from the transformation zone and cervical canal using Cervex-Brush. The results were interpreted according to the TBS classification (The Bethesda System) [10].

Genotyping of HPV DNA of high carcinogenic risk types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, and 59 isolated from clinical samples was performed using the RealBest DNA HPV genotype kit, by real-time PCR with hybridization-fluorescent detection of PCR products on a CFX-96 REAL TIME amplifier (Bio-Rad USA).

The state of the exo- and endocervix was assessed by extended colposcopy according to the standard technique using an optical colposcope Carl Zeiss E. (Germany). The findings were interpreted according to the International Classification of Colposcopic Terminology³.

Determination of serum folic acid concentration was performed once in the morning after fasting between 08:00 and 11:00 a.m. by chemiluminescent immunoassay on microparticles, Architect i2000 (Abbott), using the ADVIA Centaur FOL reagent kit manufactured by Siemens Healthcare Diagnostics (Germany). The reference values of folic acid concentration were 3.1–20.5 ng/ml.

The results were statistically processed with parametric and nonparametric analysis methods using IBM SPSS Statistics 28.0.1.1 (developed by IBM Corporation), STATISTICA 13.5.0.17 (developed by StatSoft.Inc), and MedCalc 20.027.

The intergroup differences were compared using the nonparametric Kruskal-Wallis test for independent samples with Bonferroni correction. The RANFOR (random forest) and TREE (classification trees) algorithms were used to find solving rules using SPSS. Nominal data were compared using Pearson's χ² criterion.

The relationship between the phenomena was assessed by calculating Spearman's rank correlation coefficient. ROC curve analysis was used to predict a particular disease outcome. The quality of the prognostic model was assessed based on the area under the ROC curve with standard error and the 95% confidence interval (CI) and statistical significance level.

Results

The patients' complaints at the time of consultation were represented by the comparable frequency of abnormal vaginal discharge (p>0.005, χ²=4.218), itching and burning sensations in the vagina and in the external genitalia (p>0.005, χ²=5.302), and dyspareunia (p>0.005, χ²=3.437). The proportions of patients with no complaints did not differ significantly between the study groups (p>0.005, χ²=2.371).

The age of the patients ranged from 21 to 49 years old. The distribution of patients by five-year age periods showed no significant intergroup differences (p>0.05, χ²=8.641). The proportion of women aged 21 to 24 years old was the highest in Group I (with disease regression), which was 2.1-fold higher than in Groups II and III.

Examination of social status, marital relations, and promiscuity showed no significant intergroup differences (p>0.005, χ²=2.651 and p>0.005, χ²=3.597, respectively). However, the highest proportion of unmarried patients with frequent changes of sexual partners was in Group III (with disease progression), as shown in Table 1.

The majority of the patients in the study groups were nonsmokers (p>0.005, χ²=1.027). The most frequent adherence to tobacco smoking was revealed in Group III patients (41.2% vs. 30.0% in Group II and 27.9% in Group I).

The age of menarche varied from 10 to 13 years old and did not differ significantly among the patients of all the groups under investigation (p>0.005, χ²=2.167). With the highest frequency in all clinical groups, the onset of menstrual function was between 12 and 13 years of age.

Examination of menstrual cycle parameters did not reveal any significant intergroup differences either in menstrual cycle duration, duration of menstrual bleeding and its volume (p>0.005, χ²=1.472). Various menstrual cycle disorders in the patients of the studied groups occurred with a comparable frequency (p>0.005, χ²=5.035).

The analysis of puberty onset in Table 2, which was analyzed by ages 12–14 years old, 15–17 years old, and 18 years old and older, revealed that the onset in Group I was significantly later than in Groups II and III (p<0.0001, χ²=12.82).

An analytical study of the use of means of protection against unwanted pregnancy demonstrated a consistently low rate of application of barrier methods in all of the study groups. The use of combined oral contraceptives was most frequent in Group I and II patients (60.5% and 53.3%, respectively). Intrauterine devices and interrupted intercourse (29.4%) were the preferred contraceptive methods for Group III patients (with disease progression). In the same group, the failure to use methods of protection against unwanted pregnancy occurred in 5.9% of the patients. In spite of some peculiarities of the contraceptive methods used, there were no significant differences between the examined groups (p>0.005, χ²=10.3).

Comparative analysis of obstetric history data showed that the patients of the studied groups had no significant differences in the parity of pregnancies (p>0.005, χ²=3.605) and births (p>0.005, χ²=1.841). Most women in each clinical group had a history of pregnancy (69.8%; 80.0% and 64.7%, respectively, by group), which in nearly every other patient ended in childbirth (48.8%; 50.0% and 41.5%).

Anamnestic data on previous gynecological diseases showed no significant differences in the prevalence of various nosologies (p>0.005, χ²=3.611). These data are presented in Table 3.

There were no significant differences in the history of vaginal microbiota disorders (bacterial vaginosis, aerobic vaginitis) (p>0.005, χ²=1.12) (p>0.005, χ2=5.73) among the patients in the studied groups. History of sexually transmitted diseases (chlamydia, trichomoniasis, gonococcal infection) was most common in Group II and III patients (43.3% and 52.9% vs. 25.6% in Group I). The proportion of women in Group I with a poor history of infection was the highest (p>0.05, χ²=2.965).

Somatic history was also comparably burdened with various extragenital diseases (p>0.005, χ²=2.724) and is presented in Table 4.

The laboratory examination of the state of the vaginal microbiota during the patients' treatment did not reveal any significant differences between the studied groups (p>0.005, χ²=2.305). Table 5 shows that the combined forms of nonspecific vaginal infections (bacterial vaginosis, aerobic vaginitis, candidiasis vulvovaginitis) were detected 4 times less frequently in Group I (disease regression) than in Groups II (infection persistence) and III (disease progression).

The results of the study of vaginal discharge for inflammatory response demonstrated that Group III (with disease progression) had the highest incidence of vaginal inflammatory response (4.4-fold) compared to Groups I and II. These results are shown in Table 6. A significantly higher proportion of the patients with a physiological pH of the vaginal fluid was found in the group of women with disease progression (p<0.002, χ²=3.512).

The study of the viral load, one of the criteria for stratifying patients into clinical groups, showed fundamental differences, as shown in Figure 1. After 6 months of a dynamic follow-up, Group I showed no clinically significant concentrations of HPV HCR (more than 105 genomic equivalents). The rate of patients with clinically insignificant levels of the viral load (less than 103-5 genomic equivalents of HPV HCR) was reduced 3.2-fold (90.7% of women) (p<0.003, χ²=1.341). The proportion of women with clinically insignificant viral load (less than 103 genomic equivalents of HPV HCR) increased 2.4-fold (p<0.002, χ²=1.413) (Figure 1). These patients also showed normalization of cervical smear and colposcopic findings, and there was no expression of p16INK4a protein (Figure 2).

Group II (with persistent disease) showed no significant changes in the parameters of laboratory and instrumental methods of examination (Figure 1). The results of cervical epithelial cytology (LSIL), the rate of detecting abnormal colposcopic pictures of grade I, as well as the level of the viral load and p16INK4a protein expression did not differ significantly in the dynamics of observation. Overall, the largest proportion at the time of consultation, as well as 6 months earlier, was patients with clinically insignificant and insignificant HPV HCR load (90.0% and 93.3% of patients, respectively) (Figure 1). The rate of p16ink4a oncoprotein expression detection did not differ from the previous one and was detected in 16.7% and 13.3% of patients in this group, respectively.

Group III (with disease progression) showed a significant increase (2.7-fold) in the HPV HCR viral load to the clinically significant level or higher (105 and more genomic equivalents of HPV HCR) compared to Groups I and II (p<0.001, χ²=39.192) (Figure 1). In these patients, cytological changes in the cervical epithelium and abnormal colposcopic patterns were consistent with persistent changes within LSIL. There was also a significantly higher expression of p16INK4a protein compared to comparison groups (p<0.001, χ²=24.4) (Figure 2).

The results of a comparative clinical and statistical analysis of patients with HPV-associated LSIL, as well as an evaluation of the diagnostic significance of the obtained expression data using ROC analysis, did not reveal a sufficient prognostic significance of distinctive parameters that could be used as predictors of disease outcome: regression, persistence, or progression. The area under the curve was 0.768.

Given the low prognostic potential of a number of parameters distinguishing the patients of the study groups, the authors examined the serum concentrations of vitamin B₉ (folic acid) as a possible informative parameter for the risk of cervical squamous cell intraepithelial neoplasia progression.

In Group I (with disease regression after 6 months), folic acid levels were significantly higher than in Group II (17.59±1.069 and 8.91±1.996; p=0.0003, respectively), and also in Group III (with disease progression), wherein serum levels were lowest (2.98±0.112; p= 0.0001).

Assessment of the diagnostic significance of the obtained expression data using ROC analysis demonstrated a sufficient prognostic significance of a set of parameters (early sexual debut, cytological detection of LSIL, HPV HCR viral load level ≥105 genomic equivalents of HPV HCR, expression of p16INK4a oncoprotein, the serum folic acid level less than 3.1 ng/mL) that may be considered as markers of disease progression. The area under the curve was 0.849. The comparison of the areas under the ROC curves in a comparative clinical and statistical analysis showed that when serum folic acid levels were taken into account, the prognostic significance of the disease progression prediction model was higher.

Discussion

The prevalence of LSIL ranges from 1.6% to 7.7% of the female population of reproductive age, indicating the social significance of the development and application of clear algorithms for the management of such patients. Since most cases of LSIL end in spontaneous regression, foreign and domestic clinical guidelines do not suggest an active treatment tactic [11].

According to this study, the factors significantly distinguishing patients with persistence or progression of low-grade cervical intraepithelial lesions from women with disease regression are earlier age of sexual debut, detection of low-grade atypical changes of cervical epithelium, a clinically significant level of HPV HCR viral load, and detection of p16INK4a protein expression.

At the same time, the clinical and statistical analysis of the patients did not allow answering a number of questions. If most anamnestic and follow-up data were comparable, why did the scenario of HPV-associated low-grade cervical intraepithelial lesions translate into disease progression in some patients and not in others? What mechanisms underlie the increase in viral load (105 or more genomic equivalents of HPV HCR)? What determines the more pronounced expression of p16INK4a protein, a sensitive indicator of the transition of mild potentially reversible lesions (LSIL) to severe lesions in HPV-affected cervical epithelium?

A number of numerous studies have now established that low folic acid levels are associated with an increased risk of a number of conditions (neural tube defects, cardiovascular disease, anemia) [12][13].

It has been convincingly demonstrated that folate, a folic acid-based chemical compound, is essential for nucleotide synthesis and DNA replication, which are the main metabolic processes that ensure the physiological division and normal growth of all cells in the human body [14]. The involvement of folate status in the development of some cancers is also discussed – colorectal, breast and a rapidly increasing number of its other localizations, such as lung, pancreas, stomach, esophagus, leukemia, skin, and endometrium [15].

Conclusion

The present study established the relationship of vitamin B9 concentrations with the replication level of HPV HCR, as well as with the activity of p16INK4a protein expression, which allows the authors to consider serum folate levels as an informative factor playing a significant role in the prognosis of the course of cervical squamous cell intraepithelial neoplasia of low grade. Clarification of the role of folate status in determining individual risk for HPV persistence, cervical dysplasia, and progression is an important tool for monitoring HPV-infected patients. The obtained results justify the need to select tactical options for the management of patients based on a sufficiently accurate method for predicting the progression of LSIL, which will reduce the risk of realization of HPV infection into the clinical form of the disease without violating the algorithm of current clinical guidelines.