Page 121 - Strategies for non-invasive managementof high-grade cervical intraepithelial neoplasia - prognostic biomarkers and immunotherapy Margot Maria Koeneman

Table 1. Patient characteristics
Age (mean, SD, range)
Complete follow-up (n, %)
Spontaneous regression (n, %)
HPV 16/18 negative (n, %)
Weak p16 staining (n, %)
Weak KI67 staining (n, %)
Not smoking at time of diagnosis (n, %)
Last PAP smear result before CIN2 diagnosis <PAP 3 (n, %)
No more than one CIN2 lesion in case of multiple biopsies (n, %) No use of oral contraception at time of diagnosis (n, %) Nulliparity (n, %)
Prognostic factors for spontaneous regression
-
-
-
-
-
-
13 (23%) -
-
7 (13%) 9 (16%)
37 (9, 17-61) 45 (80%) 34 (61%) 38 (68%) 36 (64%) 14 (25%) 21 (49%)
5 (9%)
51 (91%)
27 (55%) 6 20 (43%)
Prognostic factors for hrHPV positive CIN2
Predictor
Missing values
Outcome
The results of the logistic regression analysis are presented in table 2. Of all studied potential prognostic factors, only negative smoking status and nulliparity were significantly associated with disease regression (OR 3.84 [1.04-14.21] and 5.00 [1.32-19.00], respectively, in the univariable analysis). Both effects remained after correction for age and HPV16/18 in a multivariable regression analysis. With regard to age, similar conclusions were obtained if age was dichotomised (age <= 25 or age <=30). In smoking women, disease regression occurred in 10 out of 22 women (46%), compared to 16 out of 21 non-smoking women (76%). In parous women, regression occurred in 12 out of 27 women (44%), compared to 16 out of 20 nulliparous women (80%). Nulliparous women who did not smoke (n=13) showed disease regression in 11 cases (85%). Smoking, parous women (n=15) showed disease regression in five cases (33%). Logistic regression analysis could not be applied to the factor ‘last PAP smear result’, as there were only five women with a result <PAP3, all of which showed spontaneous regression. Fisher’s exact test was applied and showed a p-value of 0.143.
119