O2. HIV-testing women with cervical dysplasia

Christina Carlander1,2, Gaetano Marrone 1,3, Kristina Elfgren4, Pär Sparén 5, Anders Sönnerborg1,6
Affiliates: 1Unit of Infectious Diseases and Dermatology, Department of Medicine Huddinge, Karolinska Institutet, Stockholm, Sweden. 2Centre for Clinical Research, County Hospital Västerås, Västerås, Sweden. 3Department of Public Health Sciences, Karolinska Institutet, Stockholm, Sweden 4CLINTEC, Department of Obstetrics and Gynaecology, Karolinska University Hospital Huddinge, Stockholm, Sweden. 5Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden. 6Division of Clinical Microbiology, Department of Laboratory Medicine, Karolinska Institutet, Stockholm, Sweden.

A majority of individuals diagnosed with HIV in Sweden are late presenters, defined as CD4 count <350/mm3 or/and AIDS at time of HIV-diagnosis (Brännström 2015). When the undiagnosed HIV-prevalence in a certain population is ≥0.1% HIV-testing is suggested cost-effective and it may be cost-effective even at levels of ≥0.05% (Sanders 2005). A European multicenter study found that cervical or anal cancer/dysplasia and seven other indicator diseases fulfilled the ≥0.1% criterion for cost effectiveness (Sullivan 2013). We used Swedish national registry linkages to investigate the prevalence of undetected HIV among women diagnosed with high-grade cervical intraepithelial neoplasia (CIN2+).

All women, born between 1940 and 1990, living in the Counties of Stockholm or Gothenburg sometime between 1990 and 2014, with at least one cervical cytology test or cervical biopsy registered in the Swedish National Cervical Screening Register (NKCx) were included (n= 960 577). All women with a diagnosis of cervical intraepithelial neoplasia grade 2, grade 3, adenocarcinoma in situ or invasive cervical cancer were identified (n=62 874). By linking our study population with the national HIV register (InfCare HIV), all women with CIN2+ diagnosed before HIV-diagnosis were identified. By applying a CD4+ T-cell decline trajectory model we estimated the time of HIV seroconversion. We thus assessed the proportion of undetected prevalent HIV among women diagnosed with CIN2+ compared to women with a normal/mildly abnormal cytology/biopsy.

The prevalence of women with undetected HIV at time of CIN2+ diagnosis was 0.06% (95% CI 0.04-0.08) compared to 0.04% (95% CI 0.04-0.04) among women with normal/mildly abnormal cervical cytology/biopsy (p=0.02). Among migrants the prevalence of undetected HIV was 0.53% (95% CI 0.34-0.72) vs. 0.15 % (95% CI 0.13-0.17) respectively among women with and without CIN2+ (p< 0.01), while among women born in Sweden the prevalence was 0.02% (95% CI 0-0.03) vs. 0.02% (95% CI 0.02-0.02) respectively (p=0.43). The relative risk of being diagnosed with HIV among women with CIN2+ compared to women with normal/mildly abnormal cytology/biopsy was 1.53 (95% CI 1.09-2.14) among all women, 3.5 (95% CI 2.4-5.2) among migrants and 5.1 (CI 3.2-8.3) among women born in Sub-Saharan Africa. Women with CIN2+ prior to HIV-diagnosis had a significantly lower nadir CD4 count compared to HIV-infected women with normal/mildly abnormal cytology/biopsy (nadir CD4 count 146 vs. 210/mm3; p=<0.01).

In this population-based national register-study we found a prevalence of undetected HIV of 0.53% among migrant women diagnosed with CIN2+, suggesting that HIV-testing this population would be cost-effective. The prevalence of undetected HIV among all women, although significantly higher among those diagnosed with CIN2+, did not reach the recommended level of cost-effectiveness of ≥0.1%, but did reach the level of ≥0.05%, which has also been suggested cost-effective. Our results implies that HIV testing should be done at least in specific subpopulations of women with CIN2+ diagnosis.

We thank Pouran Almqvist at MEB, KI for help with extracting data from NKCx and Emmi Andersson, KI for help with extracting data from InfCareHIV. This study was funded by Medicines against AIDS.