$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
This retrospective cohort study was conducted at Shengjing Hospital of China Medical University and approved by the Institutional Medical Ethics Review Committee (2023PS846K). The research adhered to the principles of the Declaration of Helsinki, and written informed consent was obtained from all participants before study enrollment. The reagents and the equipment used are listed in the Table of Materials.
1. Patient selection
A total of 36 patients who underwent hysterectomy for cervical malignancy were enrolled based on predefined criteria. Inclusion criteria comprised: age 32–76 years; documented preoperative HR-HPV positivity; persistent HR-HPV positivity at the 6-month postoperative follow-up; and the absence of colposcopic or histopathological abnormalities. Exclusion criteria included: acute genital tract inflammation; renal or hepatic dysfunction; inability to comply with follow-up requirements; cardiopulmonary abnormalities; and detection of HSIL/LSIL.
2. Treatment procedure and dosing regimen
Patients allocated to the Paiteling treatment group received therapy administered in 7-day cycles, typically spanning 4–8 cycles. The therapeutic protocol involved three progressive stages tailored to lesion regression and visual inspection with acetic acid findings. In the first stage, healthcare professionals applied a local wet dressing with Paiteling solution for 15 min, three consecutive days per week, with sea buckthorn oil administered on intervening days to facilitate wound healing. This initial phase continued for 1–8 weeks until the complete resolution of visible lesions. The second stage involved the continuation of the application protocol for 3–6 weeks to consolidate therapeutic effects. Finally, the third maintenance stage involved dressings with a 2% diluted Paiteling solution applied for 10 min daily during the first month, transitioning to an every-other-day application for the subsequent 2 months. Conversely, patients in the interferon treatment group received recombinant human interferon gel administered via .a graduated vaginal applicator. The gel was applied to the cervical surface and cervical canal at a standardized dosage of 1 g per session every other day for a 3-month duration. Patients maintained recumbency for approximately 30 min post-application to optimize drug retention. For both treatment groups, follow-up evaluations were conducted three months post-treatment cessation.
3. Efficacy evaluation and follow-up
The primary efficacy endpoint was HPV clearance at the 12-month follow-up, defined as a composite of HPV-DNA <1.0 pg/mL, normal cytological findings, and the absence of colposcopic abnormalities. HPV nucleic acid detection employed the second-generation hybrid capture methodology, enabling the genotyping of 14 high-risk HPV types. Analytical thresholds defined positivity at an HPV DNA content ≥1.0 pg/mL and negativity at <1.0 pg/mL, following standard specifications. ThinPrep cytologic test samples were collected using approved disposable cervical cell collection systems during non-menstrual intervals, with strict avoidance of vaginal medication preceding sampling. Cervical secretions were cleared using a saline wipe or cotton ball before specimen collection to minimize blood and mucus contamination. Cytological abnormalities were defined as findings of atypical squamous cells of undetermined significance or greater severity. Furthermore, patients with persistent HPV positivity underwent comprehensive colposcopic evaluation using standard equipment. The examination protocol included a systematic assessment of cervical coloration, vascular architecture, and tissue characteristics. Dynamic tissue responses were documented following acetic acid application, and iodine staining patterns were recorded (Figure 1).
4. Statistical analysis
Data analysis utilized R software (version 4.5.1) with complementary packages including tidyverse, broom, survival, gtsummary, pROC, and logistf. Analytical approaches encompassed descriptive statistics, group comparisons using t-tests, Chi-square tests, or Fisher’s exact tests as appropriate, univariable and multivariable logistic regression, subgroup analysis employing Firth’s penalized likelihood method, and predictive modeling with ROC curve analysis. Statistical significance was established at p. < 0.05 (two-tailed). Due to the modest sample size, 95% confidence intervals for logistic regression models were estimated using the profile likelihood method, and corresponding p-values were derived from likelihood ratio tests.