RESUMEN
OBJECTIVE: To compare outcomes of patients with pure adenocarcinoma-in-situ (AIS) and mixed AIS/CIN 2/3 lesions including the incidence of AIS persistence, recurrence and progression to adenocarcinoma. DESIGN: Retrospective cohort study. SETTING: Statewide population in Western Australia. POPULATION: Women diagnosed with AIS between 2001 and 2012. METHODS: We conducted a retrospective, population-based cohort study. MAIN OUTCOME MEASURES: De-identified linked data were utilised to ascertain the association between patient age at excisional treatment, margin status, lesion type, lesion size, and risk of persistent AIS (defined as the presence of AIS <12 months from treatment), recurrent AIS (≥12 months post-treatment), and adenocarcinoma. RESULTS: 636 patients were eligible for analysis. The mean age was 32.3 years and median follow-up interval was 2.5 years. Within the study cohort, 266 patients (41.8%) had pure AIS and 370 (58.2%) had mixed AIS/CIN 2/3. Overall, 47 patients (7.4%) had AIS persistence/recurrence and 12 (1.9%) had adenocarcinoma. Factors associated with persistence/recurrence were pure AIS (hazard ratio (HR) 2.3; 95%CI 1.28-3.94; P = 0.005), age >30 years (HR 2.1; 95%CI 1.16-3.81; P = 0.015), positive endocervical margins (HR 5.8; 95%CI 3.05-10.92; P = <0.001) and AIS lesions >8 mm (HR 2.5; 95%CI 1.00-6.20; P = 0.049). A histologically positive AIS ectocervical margin was not associated with persistence/recurrence. CONCLUSION: In this study, pure AIS was associated with greater risk of persistence/recurrence than was mixed AIS/CIN 2/3. AIS lesions >8 mm and positive endocervical margins were significant predictors for persistent or recurrent disease. TWEETABLE ABSTRACT: Pure cervical adenocarcinoma-in-situ (AIS) may have greater risk of recurrence than AIS co-existing with CIN 2/3.