Sharpless, Kathryn E. MD, PhD; Schnatz, Peter F. DO; Mandavilli, Srinivas MD; Greene, John F. MD; Sorosky, Joel I. MD
The Bethesda system was introduced in 1988 to provide a uniform means of reporting cervical cytology findings.1 Two categories, atypical squamous cells of undetermined significance (ASC-US) and atypical glandular cells of undetermined significance (AGC-US), were introduced at that time. Now known as atypical squamous cells (ASCs) and atypical glandular cells (AGCs), these 2 findings are substantially different. The appearance of ASCs on a Pap test is a relatively common finding and is associated with a low incidence of malignancy, whereas AGCs are less common and are associated with a high rate and wide variety of premalignant and malignant diseases.2
Because of the risk of serious disease, AGCs on a Pap test require a comprehensive diagnostic evaluation.3 On the basis of current management guidelines, a minimum of a colposcopy examination with endocervical curettage (ECC) is recommended. Women aged 35 years and older and those who have other risk factors for endometrial cancer require an endometrial biopsy in addition to the standard workup. Two categories of women at high risk include those with AGC characterized as “favor neoplasia” and those with persistent AGC, which is defined as having at least 2 AGC-US Pap tests. Women with either of these clinical scenarios require a secondary evaluation consisting of a cone biopsy (loop electrosurgical excision procedure [LEEP] or cold-knife conization).
Two recent surveys found that few clinicians chose to manage women with AGC according to suggested practice guidelines.4,5 This study was designed to estimate the extent and content of diagnostic evaluations that women with AGC actually receive from their providers.
MATERIALS AND METHODS
After approval of the study protocol by the Hartford Hospital Institutional Review Board, cases of AGC-US Pap test were identified retrospectively from a pathology database during the years 1998–2001.6 During this time period, 241,224 Pap tests were performed, 648 of which were AGC-US. The 648 cases of AGC-US came from 634 women. A total of 157 women were not included in the study sample: 23 with a prior AGC-US Pap test, 7 who had a total hysterectomy, 14 with known cervical or uterine disease likely accounting for the AGC-US Pap test, 34 with concurrent ASC-US, 2 with concurrent squamous intraepithelial lesions on cytology, and 77 who lacked any follow-up data. Four hundred seventy-seven subjects, and their 477 Pap tests, formed the final study sample. The sample consisted of women with an initial AGC-US Pap test from a cervical source lacking known cervical or uterine disease or concurrent ASC-US or squamous intraepithelial lesion.
The subjects’ providers, ages, type of evaluation, and Pap test subclassification were recorded. On the basis of their descriptive modifiers (ie, “favor reactive” or “cannot exclude significant endometrial or endocervical pathology”), Pap tests were categorized into 1 of the following 3 groups: benign-appearing, malignant-appearing, or unspecified. Cytologic and histologic follow-up were evaluated for 2 years from the initial AGC-US Pap test. Disease was defined as dysplasia with a high-grade intraepithelial lesion or greater. Diagnoses made after 1 year from the initial AGC-US Pap test were considered “delayed diagnoses” in this study.
On the basis of consensus guidelines published by the American Society for Colposcopy and Cervical Pathology, the initial histologic evaluation for women younger than 35 years of age with an AGC-US Pap test includes a minimum of a colposcopy examination and ECC, and the initial histologic evaluation for women aged more than 35 years also includes an endometrial biopsy.3,7 We used this definition and assumed that women who underwent an ECC also underwent a colposcopy examination at the same time (with or without a biopsy). Those procedures were considered part of the comprehensive evaluation if they occurred from 6 months before to 12 months after the initial AGC-US Pap test. Procedures performed up to 6 months before the initial AGC-US Pap test were included to account for women who had had recent procedures that were not repeated after the AGC-US Pap test. A secondary evaluation consisted of a diagnostic cone biopsy, either a LEEP or a cold-knife conization.
Statistical analysis was performed using the χ2 test with Yates’ correction (http://www.unc.edu/∼preacher/chisq/chisq.htm). P values of .05 were considered statistically significant. Confidence intervals (CIs) of proportions (VassarStates, Poughkeepsie, NY; available at: http://faculty.vassar.edu/lowry/prop1.html) and odds ratios (ORs) and their 95% CIs (J. R. Hutchon; available at: http://www.hutchon.freeserve.co.uk/ConfidOR.htm) were calculated using online software.
Most women in the study sample were estimated to be white suburban women on the basis of their town of residence.6 The medical providers were mostly obstetrician/gynecologists (88%). Other providers included internal medicine physicians (10%), family practice physicians (1%), and physicians’ assistants and advanced practice registered nurses (1%).
The overall rate of histologic evaluation of women with an AGC-US Pap test was 64% (303/477 with a 95% CI 59–68%; Fig. 1). Women with benign-appearing AGC-US Pap test results and women with unspecified AGC-US Pap test results had similar rates of histologic evaluation (60% [209/351] and 59% [42/71], respectively). Women with malignant-appearing AGC-US Pap test results had a significantly higher rate of histologic evaluations (95% [52/55]) than both women with unspecified and benign-appearing AGC-US Pap test results (OR 12, 95% CI 3–42; P < .01 and OR 12, 95% CI 4–39; P < .01, respectively).
Overall, a comprehensive initial evaluation was performed in 36% (171/477 with a 95% CI 32–40%; Fig. 1) of the study sample. Women with malignant-appearing AGC-US Pap test results had a higher rate of comprehensive evaluations (73% [40/55]) than both women with benign-appearing (30% [104/351]) and unspecified (38% [27/71]) AGC-US Pap test results (OR 4, 95% CI 2–9; P < .01 and OR 6, 95% CI 3–12; P < .01, respectively).
The frequency of histologic evaluations was similar between age groups: 61% (67/110) in the younger-than-35-years age group and 64% (236/367) in the 35-years-or-older age group (Fig. 2). In contrast, 57% (63/110) of women younger than the age of 35 underwent a comprehensive evaluation compared with 28% (103/367) of women aged 35 year and older (OR 3.4, 95% CI 2.2–5.3; P < .01). The rates of ECC, cervical biopsy, and endometrial sampling in each age group are shown in Figure 3.
The rate of histologic evaluation among women with persistent AGC-US (76%, 22/29, with a 95% CI 56–89) was similar to the rate of histologic evaluation among women with a single AGC-US Pap test (66%, 281/448, with a 95% CI 58–67). In contrast, the rate of comprehensive evaluations was higher in women with persistent AGC-US Pap tests (66%, 19/29, with a 95% CI 46–81) than women with a single AGC-US Pap test (34%, 152/448, with a 95% CI 30–39; OR 4, 95% CI 2–8;P < .01).
Three percent (12/477) of the study sample had diagnoses made 1 year from the initial AGC-US Pap test. Those included 9 cases of high-grade squamous intraepithelial lesions and 1 case each of cervical glandular dysplasia, squamous cell carcinoma (stage 1A), and endometrial adenocarcinoma (stage 1A). The number of cases with delayed diagnoses of the total number of cases of disease was 29% (12/42). Seven of the cases of delayed diagnoses had AGC-US Pap test results that were classified as benign-appearing (58%), whereas 3 were classified as malignant-appearing (25%), and 2 were unspecified (17%). Six cases (50%) had persistent AGC-US Pap tests. Among the delayed diagnoses, the rates of histologic evaluations and comprehensive evaluations (all of which were negative) within 1 year of the initial AGC-US Pap test were both 58% (7/12). Of these 12, 3 (25%) were followed by repeat Pap test and 2 (16.7%) did not have additional follow-up until 1 year after the initial AGC-US Pap test. None of the cases of delayed diagnoses had a secondary evaluation consisting of a diagnostic cone biopsy.
Eight percent of women with persistent AGC-US Pap tests (2/26 with a 95% CI 1–27) had a secondary evaluation after an initial negative evaluation. Six of the 24 women (25%) without a secondary evaluation subsequently were diagnosed with disease. In women with a malignant-appearing AGC-US Pap test result, a secondary evaluation was performed after an initial negative evaluation in 2% (1/42 with a 95% CI 0–14). Three of the 41 women (7%) without a secondary evaluation subsequently had delayed diagnoses.
The rate of histologic evaluation in our study (64%) was similar to other large studies, which report rates of 45–88%.8–15 The low rate histologic evaluation and the reliance on follow-up Pap tests are of concern because Pap tests have reported false-negative rates of 8–50% and low sensitivity for identifying glandular lesions.16–18 Perhaps management guidelines were not followed because of a failure of clinicians to distinguish between ASC-US and AGC-US.5 Compared with women with AGC-US Pap tests, women with ASC-US Pap tests have a lower rate of malignancy (0.1%), and those with low risk can be followed up with a Pap test in 4–6 months.2
In 1998, the management of women with an AGC-US Pap test required a minimum of a colposcopy examination plus an ECC for the initial histologic evaluation.19–22 Published guidelines also recommended an endometrial biopsy for women aged 35 years or more and those with other risk factors for endometrial cancer. We found that only 36% of our study population had an appropriate and thorough evaluation (“comprehensive evaluation”).
We found that the rates of histologic and comprehensive evaluations were higher in women with AGC-US Pap test results that were subclassified as malignant-appearing than in those that were subclassified as benign-appearing or unspecified. These findings suggest that many clinicians are not aware that the subclassification system is not entirely reliable6 and that all women with an AGC-US Pap test require a comprehensive evaluation.
Our findings suggest that few women aged 35 years and older (28%) are receiving comprehensive evaluations. The low rate was mostly attributable to the low rate of endometrial biopsies (39%), in addition to an overall low rate of any histologic evaluation (64%). The low rate of endometrial biopsies in women aged 35 years and older suggests that a lack of knowledge exists regarding the risk of endometrial pathology in older women with AGC-US Pap tests.
In this study, 3% of cases of AGC-US (26% of the diagnoses) were delayed diagnoses. The delay in diagnosis in those cases can be partially explained by the low frequency of histologic and comprehensive evaluations. In women with comprehensive evaluations, the delay in diagnosis may reflect the low sensitivity of the colposcopy examination with directed biopsy for glandular lesions, which tend to be smaller and located higher in the endocervical canal,23,24 and the high false-negative rate of ECC.25 In this study, 67% (8/12) of the women with delayed diagnoses had a negative comprehensive evaluation (8/12). The incidence of false negatives among women with a comprehensive evaluation was 1% (8/145).
In 1998, cone biopsy (cold-knife or LEEP) of the cervix was the management recommendation for women with a malignant-appearing AGC-US Pap test with a negative initial evaluation and for women with persistent unexplained AGC-US Pap tests.20–22 Among the cases of delayed diagnoses in this study, 8 cases (67%) were either persistent AGC-US or were subclassified as malignant. None of the cases of delayed diagnosis had a secondary evaluation. These findings suggest a lack of awareness of the high risk in those women.
Among high-risk subjects, 2% of women with malignant-appearing Pap tests and 8% of women with persistent AGC-US Pap tests had a cone biopsy after an initial negative evaluation. These findings concur with 2 recent surveys of physician management, which found that few physicians choose to manage persistent AGC-US with surgical excision after a negative initial evaluation.4,5
This was a retrospective study based on information from a pathology computer database, which does not include clinical information or provide a means of determining the reasoning behind management decisions. Some patients may not have undergone a comprehensive evaluation because benign disease was identified accounting for the AGC-US Pap test. Practice guidelines suggest that all patients with AGC-US require a comprehensive evaluation. However, it could be argued that patients with benign diagnoses accounting for the AGC-US need no further evaluation. In that case, the number of women with AGC-US with an appropriate evaluation in this study would be higher. Additionally, the percent with late diagnoses may be larger than we estimated because evaluations were often incomplete. For this study, we had no direct means of identifying women who were referred for and underwent a colposcopy examination. In the absence of a punch biopsy specimen, we assumed a colposcopy examination was performed if an ECC was performed because this is standard practice in our institution. The rates of comprehensive evaluations, therefore, would be even lower if there were women who underwent an ECC without a colposcopy examination.
The purpose of this study was to assess whether physicians have an appropriate appreciation for AGC-US. Our data suggest that women with AGC-US Pap tests are undermanaged in the initial evaluation and even more so in the secondary evaluation. Less than 40% of women had a thorough initial evaluation. Additionally, less than 10% of women with a malignant-appearing Pap test result or persistent AGC-US Pap test results had a secondary evaluation after a negative initial evaluation. Twenty-nine percent of the diagnoses were delayed, occurring more than 1 year after the initial AGC-US Pap.
Our findings suggest that physicians associated with a university-affiliated hospital, most of whom were obstetricians and gynecologists, failed to follow practice guidelines for the management of women with AGC-US Pap tests. These findings are likely representative of the clinical practices of physicians across the country because providers practicing in association with Hartford Hospital have diverse training backgrounds, and the findings concur with nationwide surveys of physicians’ practices.4,5 To increase the frequency of histologic evaluation of patients with AGC-US Pap tests, cytopathologists could recommend comprehensive histologic follow-up of all cases of AGC-US, but this may not be acceptable for many reasons, including the legal implications. Another means of increasing physician compliance with practice guidelines is continued education and testing. This may be an area that the American Board of Obstetricians & Gynecologists, as well as other primary care fields, should emphasis in their recertification materials and review articles. Education may be responsible for the improvements we observed in the quality of management during the time period of the study. Between the first 2 years and last 2 years of the study, the rate of comprehensive evaluations improved from 29% to 46%, and the rate of endometrial biopsies increased from 31% to 52%. In 2001, the new Bethesda system removed the misleading phrase “of undetermined significance,” and the American Society for Colposcopy and Cervical Pathology published consensus guidelines for the management of women with AGC.7,26 Thus, although our data suggest that few clinicians followed practice guidelines for the management of women with AGC-US Pap tests during 1998–2001, future studies may find that physicians have become more aware of the clinical significance of AGC and follow recommended management guidelines.
1. The 1988 Bethesda System for Reporting Cervical/Vaginal Cytological Diagnoses. National Cancer Institute Workshop. JAMA 1989;262:931–4.
2. Bernard-Pearl L, Smith-McCune K. Controversies in the management of ASCUS and AGCUS: two very different beasts. Curr Probl Obstet Gynecol Fertil 2001;24:7–23.
3. Levine L, Lucci JA III, Van Dinh T. Atypical glandular cells: new Bethesda Terminology and Management Guidelines [review]. Obstet Gynecol Surv 2003;58:399–406.
4. Noller KL, Bettes B, Zinberg S, Schulkin J. Cervical cytology screening practices among obstetrician–gynecologists. Obstet Gynecol 2003;102:259–65.
5. Smith-McCune K, Mancuso V, Contant T, Jackson R. Management of women with atypical Papanicolaou tests of undetermined significance by board-certified gynecologists: discrepancies with published guidelines. J Obstet Gynecol 2001;185:551–6.
6. Sharpless KE, Schnatz PF, Mandavilli S, Greene JF, Sorosky JI. Dysplasia associated with atypical glandular cells on cervical cytology. Obstet Gynecol 2005;105:494–500.
7. Wright TC Jr, Cox JT, Massad LS, Twiggs LB, Wilkinson EJ, for the ASCCP-Sponsored Consensus Conference. 2001 consensus guidelines for the management of women with cervical cytological abnormalities. JAMA 2002;287:2120–9.
8. Raab SS, Isacson C, Layfield LJ, Lenel JC, Slagel DD, Thomas PA. Atypical glandular cells of undetermined significance: cytologic criteria to separate clinically significant from benign lesions. Am J Clin Pathol 1995;104:574–82.
9. Eddy GL, Strumpf KB, Wojtowycz MA, Piraino PS, Mazur MT. Biopsy findings in five hundred thirty-one patients with atypical glandular cells of uncertain significance as defined by the Bethesda system. Am J Obstet Gynecol 1997;177:1188–95.
10. Bennett BB, Takezawa K, Wilkinson EJ, Drew PA, Hardt NS. Atypical glandular cells of undetermined significance and other glandular cell abnormalities in a high-risk population. J Lower Gen Tract Dis 1998;2:132–5.
11. Veljovich DS, Stoler MH, Andersen WA, Covell JL, Rice LW. Atypical glandular cells of undetermined significance: a five-year retrospective histopathologic study. Am J Obstet Gynecol 1998;179:382–90.
12. Chhieng DC, Elgert P, Cangiarella JF, Cohen JM. Variation in the incidence of AGUS between different patient populations. Acta Cytol 2001;45:287–93.
13. Geier CS, Wilson M, Creasman W. Clinical evaluation of atypical glandular cells of undetermined significance. Am J Obstet Gynecol 2001;184:64–9.
14. Koonings PP, Price JH. Evaluation of atypical glandular cells of undetermined significance: is age important? Am J Obstet Gynecol 2001;184:1457–61; discussion 1459-61.
15. Nasuti JF, Fleisher SR, Gupta PK. Atypical glandular cells of undetermined significance (AGUS): clinical considerations and cytohistologic correlation. Diagn Cytopathol 2002;26:186–90.
16. Jones BA, Novis DA. Cervical biopsy-cytology correlation: a College of American Pathologists Q-probes study of 22,439 correlations in 348 laboratories. Arch Pathol Lab Med 1996;120:523–31.
17. van Aspert-van Erp AJ, Smedts FM, Vooijs GP. Severe cervical glandular cell lesions and severe cervical combined lesions: predictive value of the Papanicolaou smear. Cancer 2004;102:210–7.
18. Krane JF, Granter SR, Trask CE, Hogan CL, Lee KR. Papanicolaou smear sensitivity for the detection of adenocarcinoma of the cervix. Cancer 2001;93:8–15.
19. Cervical cytology: evaluation and management of abnormalities. ACOG Technical Bulletin No. 183. Washington, DC: ACOG; 1993.
20. Kurman RJ, Henson DE, Herbst AL, Noller KL, Schiffman MH. Interim guidelines for management of abnormal cervical cytology The 1992 National Cancer Institute Workshop. JAMA 1994;271:1866–9.
21. Cox JT. ASCCP practice guidelines: management of glandular abnormalities in the cervical smear. J Lower Gen Tract Dis 1997;1:41–5.
22. Solomon D, Frable WJ, Vooijs GP, Wilbur DC, Amma NS, Collins RJ, et al. ASCUS and AGUS criteria. International Academy of Cytology Task Force summary. Diagnostic cytology towards the 21st century: an international expert conference and tutorial [review]. Acta Cytol 1998;42:16–24.
23. Selim MA, So-Bosita JL, Blair OM, Little BA. Cervical biopsy versus conization. Obstet Gynecol 1973;41:177–182.
24. Ferris DG, Krumholz BA, Jester DM, Crosby JH, Hanly MG, Messing MJ. Atypical glandular cells of undetermined significance and adenocarcinoma in situ: summoning colposcopic expertise? J Fam Pract 1996;43:181–7.
25. Denehy TR, Gregori CA, Breen JL. Endocervical curettage, cone margins, and residual adenocarcinoma in situ of the cervix. Obstet Gynecol 1997;90:1–6.
26. Solomon D, Davey D, Kurman R, Moriarty A, O'Connor D, Prey M, et al, Forum Group Members; Bethesda 2001 Workshop. The 2001 Bethesda system: terminology for reporting results of cervical cytology. JAMA 2002;287:2114–9.