Cytopathology in Focus: More aggressive follow-up for patients with AGC?

Renee L. Betancourt, MD; Avani A. Pendse, MD, PhD; Leslie Dodd, MD

May 2017—The most recent edition of the Bethesda System for Reporting Cervical Cytology classifies glandular cell abnormalities into the following broad categories: atypical (specify favored site of origin), atypical (favor neoplastic), endocervical adenocarcinoma in situ (AIS), and adenocarcinoma.¹ Generic terminology of “atypical glandular cells (AGC)” may be used if the origin of the cells cannot be determined with certainty. Nevertheless, the Bethesda System encourages pathologists and cytotechnologists to report the favored site of origin (endometrial versus endocervical) whenever possible.

According to the Bethesda System, atypical endocervical cells are defined as “endocervical-type cells that display nuclear atypia that exceeds obvious reactive or reported reactive changes, but lack unequivocal features of endocervical adenocarcinoma in situ or invasive adenocarcinoma.”¹ Atypical glandular cells is an uncommon cytologic diagnosis with an estimated frequency of less than one percent of cervical Pap test results.² Despite its relative rarity as a cytologic diagnosis, AGC is associated with a range of common benign to malignant lesions, including both squamous cell carcinoma and adenocarcinoma.

Since its inception, the main goal of the Pap test (formerly known as Pap smear) was to efficiently screen patients in order to detect cervical lesions, specifically for squamous cell carcinoma and its known precursors. While Pap testing can also detect endocervical and endometrial glandular pathology, these lesions are inherently less accessible and more challenging to diagnose than squamous lesions.³ Consequently, in many developed countries, the incidence of cervical adenocarcinoma has not decreased to the same extent that invasive cervical squamous cell carcinoma has decreased. In addition, a study published in the United States in 2000 showed that the proportion of adenocarcinomas relative to squamous cell carcinomas and to all cervical cancers had doubled, suggesting that the screening method and guidelines were inadequate at detecting precursor lesions to cervical adenocarcinoma.⁴ Moreover, a study conducted in England showed that the proportion of cervical squamous carcinomas decreased from 82.6 percent in 1989 to 70.4 percent in 2009.⁵ Conversely, the proportion of cervical adenocarcinomas increased from 13.2 percent in 1989 to 22.1 percent in 2009.⁵ Despite the shifts in incidence of individual histomorphologic subtypes, invasive cervical cancer continues to be a persistent public health concern, with 12,990 estimated new cases in the U.S. in 2016 expected to result in approximately 4,120 deaths.⁶

A national population-based cohort study published by Wang, et al.,⁷ is reported to be the first investigational study to show an association between the detection of AGC and a long-term increased risk of developing adenocarcinoma. The study design used the Swedish National Cervical Screening Registry (consisting of cytology screening and histological results dating back to 1980), as well as the Swedish National Cancer Register. The authors examined the records of 3,054,328 women ages 23–59 who had cervical cytology screening in Sweden from Jan. 1, 1980 to July 1, 2011. A total of 14,625 patients were diagnosed with AGC, of whom 330 were diagnosed with cancer (198 within six months and an additional 132 within 15.5 years). This study also found that the presence of AGC implied a significantly increased cervical cancer risk for at least 15.5 years after the initial screening detection. Aiming to extrapolate beyond the currently available evidence, the authors also developed a stratification of patients by age at first detection of AGC and by histopathological characteristics of their subsequent invasive cervical cancer.

The incidence rate ratios of adenocarcinoma after AGC during the first screening was 61-fold when compared with women who had normal cervical cytology. Not surprisingly, the majority of subsequent cervical cancer diagnoses following the detection of AGC were adenocarcinoma (74 percent), whereas most following high-grade or low-grade squamous intraepithelial lesions were squamous cell carcinoma (more than 80 percent). Of note, one limitation the authors acknowledged was that AGC was not subdivided into the separate categories proposed by the Bethesda System, and thus no conclusions can be drawn about the risk of cancer after the diagnosis of a specific subcategory of AGC.

The highest prevalence of cervical cancer at the time of initial AGC diagnosis was found in women ages 30–39, both for adenocarcinoma and squamous cell carcinoma. The desire to preserve fertility was thought to influence a more conservative treatment course following the initial AGC diagnosis in this age group. Conversely, the highest prevalence of cancers diagnosed following HSIL were found in the 50–59 age group. A diagnosis of LSIL with prevalent cancer was rare (0.2 percent), and there was no meaningful association with age. The proportion of women who underwent clinical and histological assessment after AGC was significantly lower than those with a diagnosis of HSIL but higher than LSIL. The method and frequency with which a patient was managed clinically after a cytological abnormality therefore appeared to have a direct impact on the subsequent cancer risk.

Practice guidelines for optimal management of AGC were not issued in Sweden until 2010. These guidelines recommend immediate histologic assessment of AGC, similar to the existing follow-up guidelines for patients diagnosed with HSIL. However, the authors conclude that this single, initial assessment is likely to be inadequate. Rather, there should be an additional histologic exam in one year, as well as long-term surveillance to detect precursor lesions.

When histologic assessment was performed within six months on women with AGC, there was a 2.8 percent cancer risk, which is close to that of HSIL at 3.2 percent. This highlights the need for immediate diagnostic investigation—a practice that was not in place until recent years for AGC cases. As such, the possibility exists that there is a falsely elevated incidence of both squamous cell carcinoma and adenocarcinoma diagnoses after a diagnosis of AGC as compared with HSIL cases, which tended to have a higher prevalence.

The controversial role of primary human papillomavirus testing as a potential screening tool is discussed briefly in this study, as HPV infection has been shown to also have a relation to high-risk glandular precursor lesions. The authors suggest that the initial viral screen would be coupled with cytology triage, in hopes of identifying the highest-risk AGC group for further clinical assessment or treatment, if indicated.

The data from this study illustrate the highest prevalence of cancers following a diagnosis of AGC occurs in women ages 30–39, a population that may be averse to disease management options with potential to affect future pregnancy outcomes. A recent comparison of early cervical carcinoma (FIGO stage IB–IIA) after radical hysterectomy showed that adenocarcinoma may have a worse survival outcome than squamous cell carcinoma.⁸ This further supports the concern that current management of AGC may be suboptimal when compared with that of squamous lesions. Additional investigations should be aimed at developing effective yet safe methods for evaluating patients after a diagnosis of AGC.⁹

In conclusion, the authors of this study found that AGC is associated with a high and persistent risk of cervical cancer for up to 15 years, especially in women with AGC at ages 30–39. In light of the evolving clinical significance of an AGC diagnosis, this cohort of women should undergo immediate diagnostic investigation and be followed more closely. While some may argue that this might result in overtreatment, there is compelling evidence to warrant more aggressive follow-up.

1. Nayar R, Wilbur DC, eds. The Bethesda System for Reporting Cervical Cytology: Definitions, Criteria, and Explanatory Notes. 3rd ed. Cham, Switzerland: Springer; 2015.
2. Schnatz PF, Guile M, O’Sullivan DM, Sorosky JI. Clinical significance of atypical glandular cells on cervical cytology. Obstet Gynecol. 2006;107(3):701–708.
3. Nayar R, Wilbur DC. The Pap test and Bethesda 2014. Acta Cytol. 2015;59(2):121–132.
4. Smith HO, Tiffany MF, Qualls CR, Key CR. The rising incidence of adenocarcinoma relative to squamous cell carcinoma of the uterine cervix in the United States—a 24-year population-based study. Gynecol Oncol. 2000;78(2):97–105.
5. Castanon A, Landy R, Sasieni P. Is cervical screening preventing adenocarcinoma and adenosquamous carcinoma of the cervix? Int J Cancer. 2016;139(5):1040–1045.
6. Cancer stat facts: cervix uteri cancer. National Cancer Institute Surveillance, Epidemiology, and End Results Program website. https://seer.cancer.gov/statfacts/html/cervix.html. Accessed Dec. 5, 2016.
7. Wang J, Andrae B, Sundström K, et al. Risk of invasive cervical cancer after atypical glandular cells in cervical screening: nationwide cohort study. BMJ. 2016;352:i276.
8. Lee YY, Choi CH, Kim TJ, et al. A comparison of pure adenocarcinoma and squamous cell carcinoma of the cervix after radical hysterectomy in stage IB-IIA. Gynecol Oncol. 2011;120(3):439–443.
9. Burki TK. Atypical glandular cells and risk of cervical cancer. Lancet Oncol. 2016;17(3):e96.

[hr]

Dr. Betancourt is a resident in pathology, Dr. Pendse is a fellow in surgical pathology, and Dr. Dodd is a professor and director of cytopathology—all in the Department of Pathology and Laboratory Medicine, University of North Carolina School of Medicine, Chapel Hill. Dr. Dodd is a member of the CAP Cytopathology Committee.