Results
Two hundred seventeen patients were recruited in the study including 57 LSIL(VaIN1) and 160 HSIL(VaIN2,3) patients. The general information of the patients is shown in Table 1. There was no significant difference in age, menopause status, gravidity, parity, abortion, condemn use status, cervical contact bleeding or cytologic results before laser between the LSIL(VaIN1) and HSIL(VaIN2,3) groups. There was a significant difference in HR-HPV test results before laser ablation between the two groups, and the HSIL(VaIN2,3) group showed a higher HR-HPV infection rate than the LSIL(VaIN1) group (89.4% vs 78.9%, p =0.047).
Information about concomitant or previous diseases of CIN or cervical cancer (CC) and the corresponding previous treatments are listed in Table 2. Of all the patients, 61.9% had the illness of CIN/CC.
The characteristics of VaIN lesions are listed in Table 3. The average area of VaIN lesions was 1.17cm2. Areas of 63.1% of lesions were less than 1cm2. The areas of lesions in the HSIL(VaIN2,3) group were larger than that in the LSIL(VaIN1) group with a significant difference (1.31cm2 vs 0.79cm2, p =0.055). A history of CIN/VIN in the LSIL(VaIN1) group was more common than that in the HSIL(VaIN2,3) group with a significant difference (75.4% vs 58.1%, p =0.020).
The average laser number for all the patients was 2.94. The average number of therapeutic laser ablation was 2.32 and that of strengthened laser ablation was 0.62 (Fig. 1). There was a significant difference in the total number, therapeutic number and strengthened number of laser ablation between the HSIL(VaIN2,3) and LSIL(VaIN1) groups (3.15 vs 2.35,p =0.000; 2.47 vs 1.91, p =0.002; 0.68 vs 0.44, p=0.060; respectively).
Seventeen patients lost follow-up for the assessment of pathological cure and 22 patients lost for the assessment of HPV clearance. The pathological cure rate increased along with increasing laser times in all the patients (Fig. S1). The cumulative pathological cure rate was 93.0% (186/200) and the cumulative HR-HPV negative rate was 76.9% (150/195). Multivariate logistic regression analysis showed that larger areas of VaIN lesions (p =0.048) and history of hysterectomy (p =0.017) were independent risk factors for the pathological persistence (Table S1), and that menopause (p =0.042) and non-use of condemns (p =0.068) were independent risk factors for persistent HR-HPV infection (Table S2).
In the HSIL(VaIN2,3) group, the cumulative pathological cure rate was 92.0% (138/150) and the cumulative HR-HPV negative rate was 77.5% (110/142). Three to five times of laser ablation could obtain a relatively high pathological cure rate (90.4~92.0%) and HR-HPV negative rate (68.5~77.5%) (Fig. S2). Multivariate logistic regression analysis showed that larger areas of VaIN lesions (p =0.083) and history of hysterectomy (p=0.037) were independent risk factors for pathological persistence, and that menopause (p =0.006) and immunosuppression (p =0.095) were risk factors for persistent HR-HPV infection (Table S3 and Table S4).
In the LSIL(VaIN1) group, the cumulative pathological cure rate was 96.0% (48/50) and the cumulative HR-HPV negative rate was 75.5% (40/53). Two to three times of laser ablation could obtain a relatively high pathological cure rate (92.0~93.5%) and HR-HPV negative rate (65.5~73.6%) (Fig. S3). Multivariate logistic regression analysis showed there was no risk factor for pathological persistence, and that condemn use (p =0.002) was a protective factor against HR-HPV infection (Table S5 and Table S6).
The main complication of laser ablation for VaIN was the vaginal adhesions (21/217, 9.68%). There was another case of surgical site bleeding. The complication incidence was significantly higher in the HSIL group than that in the LSIL group (13.13% vs 1.75%, p=0.014). Bivariate correlation analysis (Table S7) showed that complication incidence was related to vaginitis (p =0.004), VaIN grade (p =0.014) and laser times (p =0.046). Multivariate logistic regression analysis showed that the upper three factors were independent risk factors for complication incidence (p =0.015,p =0.029 and p =0.046, respectively).
There were 455 published studies searched out in the meta-analysis and 21 eligible articles [14, 18~37] were chosen for analysis. The overall combined effect was 0.78 (95%CI 0.72~0.84, P<0.05) for the assessment of pathological cure of laser ablation for VaIN (Fig. S4). The combined effect was 0.84 (95%CI 0.75~0.92, P<0.05) of laser ablation for both the HSIL(VaIN2,3) and LSIL(VaIN1) patients.