Outcomes in cervical screening using various cytology technologies

Abstract

Unlike for human papillomavirus screening, little is known about the possible age-dependent variation in the outcomes of cervical cytology screening. The aim of our study was to describe age-related outcomes of five cytological technologies in a population-based screening program targeting women aged 23-59 years. All cervical cytology from women residing in Copenhagen has been analyzed in the laboratory of the Department of Pathology, Hvidovre University Hospital. We studied five technology phases: (1) conventional cytology with manual reading, (2) conventional cytology with 50% automatically signed out as normal, (3) liquid-based cytology (LBC) with 50% automatically signed out as normal, (4) LBC with 25% automatically signed out as normal, and (5) LBC with 25% automatically signed out as normal and with 16 preselected areas for attention in manual reading. We calculated proportion of samples with atypical squamous cells of undetermined significance or worse (≥ASCUS) by age and technology phase. We included 391 140 samples. The proportion of ≥ASCUS increased steadily from 3.8% in phase 1 to 6.0% in phase 5. This pattern varied considerably across age groups. In women aged 23-34 years, the proportion almost doubled, relative proportion 1.96 (95% confidence interval: 1.84-2.08). An opposite development was seen in women aged 45-59 years, relative proportion 0.68 (95% confidence interval: 0.57-0.82). Technological upgrading of cytology strongly affected the outcome of cervical screening for young women. If corroborated with data from other laboratories, these findings call for caution in implementation of new cytology screening technologies.

Trends in the incidence of cervical cancer and severe precancerous lesions in Denmark, 1997–2012

“…During the period, several Danish pathology departments changed from conventional, manually read Papanicolaou (Pap) smears to automation-assisted reading of liquid-based cytology, and possibly this transition may have influenced the incidence. Two Danish studies of the impact of cervical screening technology at different hospitals in the greater Copenhagen area showed notable differences in the results obtained with the two techniques [46,47] although a metaanalysis did not find that liquid-based cytology was more sensitive or more specific for detecting cervical lesions [48]. In a study of almost 400,000 cervical samples, Barken et al [46] found that a change from conventionally read Pap smears to automation-assisted reading of liquid-based cytology increased the proportion of samples with atypical squamous cells of undetermined significance (ASCUS) or worse from 3.8 to 6.0 % during 1998-2007.…”

“…Two Danish studies of the impact of cervical screening technology at different hospitals in the greater Copenhagen area showed notable differences in the results obtained with the two techniques [46,47] although a metaanalysis did not find that liquid-based cytology was more sensitive or more specific for detecting cervical lesions [48]. In a study of almost 400,000 cervical samples, Barken et al [46] found that a change from conventionally read Pap smears to automation-assisted reading of liquid-based cytology increased the proportion of samples with atypical squamous cells of undetermined significance (ASCUS) or worse from 3.8 to 6.0 % during 1998-2007. Rask et al [47] found that use of imaging-assisted reading increased the proportion of samples with ASCUS or worse by about 30 % in all age groups.…”

Impact of technology on cytology outcome in cervical cancer screening of young and older women

“…Data from the Department of Pathology, Hvidovre University Hospital, are included in Figure . A detailed description of the technological phases in this laboratory was reported in a previous publication …”

“…As we wanted to investigate whether the trends observed in the analysis based on data from Hvidovre laboratory could be found also in other laboratories, we used the same observation period from January 1, 1998 to December 31, 2007, and included samples from women aged 23–59 years, the target age of the organized screening program until 2007. However, because of a national administrative reorganization, the catchment areas for the laboratories in Hillerød and Roskilde changed in 2007, and for these two laboratories comparable data were only available until December 31, 2006.…”

“…A recent Danish observational study, based on 390,000 cytological samples from the Department of Pathology, Hvidovre University Hospital, observed a doubling in the proportions of abnormal samples after manually read conventional cytology (CC) was replaced with modern cytological technologies, i.e ., SurePath‐based LBC and automation‐assisted reading over a period of 10 years, 1998–2007. When LBC was implemented in 2002, the impact on abnormality proportions differed substantially between the age groups, with a 14% decrease in the older women aged 45–59 years simultaneously with a 44% increase in the younger women aged 23–30 years (Fig.…”