Cytologic and biopsy findings leading to conization in adenocarcinoma in situ of the cervix

Cytologic and Biopsy Findings Leading to Conization in Adenocarcinoma In Situ of the Cervix Christine H. Shin, MD, John O. Schorge, MD, Kenneth R. Lee, MD, and Ellen E. Sheets, MD OBJECTIVE: To investigate the utility of currently available screening tests in preoperatively detecting adenocarcinoma in situ of the cervix. METHODS: Patients with a cone biopsy diagnosis of adenocarcinoma in situ from 1987 to 2000 at our institution were identified. Results from Papanicolaou smears, cervical biopsies, and endocervical curettages preceding the diagnostic cone biopsy were collected from medical records and referring providers. Fisher exact test (two-tail) was used for statistical analysis. RESULTS: The preoperative screening results preceding a cone biopsy containing adenocarcinoma in situ were available in 118 patients. Among 94 Papanicolaou smears, 65 (69%) glandular lesions and 29 (31%) squamous or unspecified lesions were reported. Biopsy and/or endocervical curettage after the 29 squamous or unspecified lesions on Papanicolaou smear detected 15 additional glandular lesions, totaling 80 (85%) of 94 cases of glandular disease detected before conization. Among all 118 cases with some form of preoperative data available, glandular disease was predicted in 100 cases (85%). In cases of suspected glandular disease, 86% were treated with cold knife cone compared with 22% in cases of suspected squamous abnormalities (P < .001). CONCLUSION: The sensitivity of detecting a glandular abnormality before a cone biopsy containing adenocarcinoma in situ is 69% with the Papanicolaou smear and 85% with the addition of biopsy and endocervical curettage. This underscores the importance of using preoperative assessment to appropriately plan treatment for a suspected glandular lesion. (Obstet Gynecol 2002;100:271– 6. © 2002 by The American College of Obstetricians and Gynecologists.)

Adenocarcinoma of the cervix is increasing in incidence and currently accounts for approximately 20 –25% of primary cervical cancers.1 The ability to detect its precursor, adenocarcinoma in situ, has important diagnostic and treatment implications.2 The average age of patients with clinically detected adenocarcinoma in situ is apFrom the Division of Gynecologic Oncology and Division of Women’s and Perinatal Pathology, Brigham and Women’s Hospital, Boston, Massachusetts.

proximately 5 years younger than those with early invasion, supporting the potential for Papanicolaou smear screening to prevent this disease.3 Cervical cytologic screening has decreased the incidence of invasive squamous cell carcinomas of the cervix by detecting preinvasive lesions.2 Colposcopically directed biopsies and endocervical curettage (ECC) have known utility in the management of squamous dysplasia.4,5 We sought to determine the effectiveness of the Papanicolaou smear, cervical biopsy, and ECC in the diagnosis of glandular lesions of the cervix by retrospectively reviewing the cytopathology and histopathology results that preceded cone biopsies containing adenocarcinoma in situ. MATERIALS AND METHODS A retrospective Institutional Review Board-approved study was conducted to identify all women diagnosed with adenocarcinoma in situ of the cervix in a cone biopsy specimen at Brigham and Women’s Hospital between December 1987 and March 2000. Patients with a coexisting squamous intraepithelial lesion (SIL) were included; patients with a coexisting invasive adenocarcinoma were excluded. Gynecologic pathologists at Brigham and Women’s Hospital had confirmed the adenocarcinoma in situ diagnosis on all cone biopsy pathology slides, some of which were prepared at other locations. However, not all preconization slides were rereviewed. The diagnosis of adenocarcinoma in situ was based on previously published pathologic criteria.6 – 8 Results of Papanicolaou smears, biopsies, and ECCs performed within 1 year of a cone biopsy containing adenocarcinoma in situ were abstracted from patient records and by correspondence with referring providers. Results were correlated with cone biopsy diagnosis of either pure adenocarcinoma in situ or mixed lesions containing a concurrent SIL. Further correlation was made between preoperative screening results and initial treatment choices. Fisher exact test (two-tail) was used for statistical analysis.

VOL. 100, NO. 2, AUGUST 2002 © 2002 by The American College of Obstetricians and Gynecologists. Published by Elsevier Science Inc.

0029-7844/02/$22.00 PII S0029-7844(02)02044-6

271

Table 1. Patient Demographics (n ⫽ 118) Median age (y) Age range (y) Baseline parity (%) Nulliparous One child ⬎One child Unknown Oral contraceptive use (%) Smoking history (%)

31 17–72 62 15 20 3 33 28

RESULTS A total of 118 (89%) of 133 women with adenocarcinoma in situ in a conization specimen had preconization cytology and/or histology data available for review. Age, parity, oral contraceptive use, and smoking history are reported in Table 1. In 94 cases, data were available from a Papanicolaou smear and colposcopically directed biopsy and/or ECC. For the remaining 24 patients, Papanicolaou smear data could not be obtained, but preconization histologic data was available from either biopsy and/or ECC. Figure 1 illustrates the preconization findings of glandular abnormalities from cytologic and histologic testing. The cytologic interpretations on Papanicolaou smears leading to the diagnostic cone biopsies are shown in Table 2. Glandular abnormalities were detected in 65 (69%) of 94 preconization Papanicolaou smears; 37 (57%) of these 65 cases were reported either as “adeno-

Table 2. Cytologic Findings on Papanicolaou Smear Preceding Diagnostic Cone Biopsy With Adenocarcinoma In Situ (n ⫽ 94) Glandular lesions, n (%) AIS AGUS, suggest AIS Adenocarcinoma AGUS, suggest adenocarcinoma AGUS, NOS Glandular atypia Squamous or unspecified lesions, n (%) HSIL LSIL ASCUS Atypical cells, suggest carcinoma Atypical cells, NOS

65 (69) 20 17 3 5 14 6 29 (31) 19 1 4 2 3

AIS ⫽ adenocarcinoma in situ; AGUS ⫽ atypical glandular cells of undetermined significance; NOS ⫽ not otherwise specified; HSIL ⫽ high-grade squamous intraepithelial lesion; LSIL ⫽ low-grade squamous intraepithelial lesion; ASCUS ⫽ atypical squamous cells of undertermined significance.

carcinoma in situ” or “atypical glandular cells suggesting adenocarcinoma in situ.” Abnormal Papanicolaou smear findings were followed by colposcopically directed biopsy and/or ECC in 79 (84%) of the 94 cases. Among the 27 patients who underwent colposcopy because of a suspected squamous or unspecified lesion on Papanicolaou smear, 15 additional glandular abnormalities were detected on biopsy and/or ECC (Figure 1). By combining Papanicolaou

Figure 1. Pathologic findings of glandular lesions on Papanicolaou smear, colposcopically directed biopsy, and/or endocervical curettage preceding diagnostic conization containing adenocarcinoma in situ. AIS ⫽ adenocarcinoma in situ; BX ⫽ biopsy; ECC ⫽ endocervical curettage. Shin. Preconization Findings of Adenocarcinoma In Situ. Obstet Gynecol 2002.

272

Shin et al

Preconization Findings of Adenocarcinoma In Situ

OBSTETRICS & GYNECOLOGY

Table 3. Correlation of Preconization Findings of a Glandular Abnormality With Cone Biopsy Findings of Either Pure Adenocarcinoma In Situ or a Mixed Adenocarcinoma In Situ/Squamous Intraepithelial Lesion Cone biopsy histology Papanicolaou, biopsy, ECC results Among patients with Papanicolaou smear,* n Glandular abnormality, n (%) Squamous abnormality, n (%) Among all patients,† n Glandular abnormality, n (%) Squamous abnormality, n (%)

94 80 (85) 14 (15) 118 100 (85) 18 (15)

Pure AIS

Mixed AIS/SIL

55 52 (65) 3 (21) 68 63 (63) 5 (28)

39 28 (35) 11 (79) 50 37 (37) 13 (72)

ECC ⫽ endocervical curettage; other abbreviations as in Table 2. * P ⫽ .003, Fisher exact test, two-tailed. † P ⫽ .001, Fisher exact test, two-tailed.

smear (n ⫽ 65) with biopsy and/or ECC (n ⫽ 15), a glandular abnormality was suggested in 80 (85%) of 94 cases before conization. The remaining 14 patients underwent treatment for presumed squamous abnormalities and were found incidentally to have adenocarcinoma in situ on the cone biopsy. Eleven (79%) of these contained a coexistent high-grade squamous lesion, whereas only 28 (35%) of the remaining 80 patients had mixed lesion on the cone biopsy (Table 3). The remaining 24 patients had data available only from biopsy and/or ECC, but presumably had an abnormal Papanicolaou smear leading to further colposcopic testing (Figure 1). Among a total of 118 cases for which preconization information was available from at least one modality (Papanicolaou, biopsy, and/or ECC), 68 cone biopsies contained pure adenocarcinoma in situ lesions and 50 contained adenocarcinoma in situ in association with SIL. Overall, 85% of adenocarcinoma in situ lesions present in conization specimens were detected preoperatively by combining histology with cytology (Table 3). The choice of initial treatment varied depending on preoperative cytologic and histologic findings (Table 4).

A preconization finding of a glandular or mixed (glandular and squamous) abnormality, whether by Papanicolaou, biopsy/ECC, or a combination of these, led to initial treatment with a cold knife cone biopsy in 86% and 74% of cases, respectively. In contrast, preconization findings of a squamous abnormality alone led to a cold knife cone biopsy in only 22% of cases; the majority were treated with loop electrical excisional procedure. DISCUSSION The proportion of adenocarcinomas of the cervix in the United States has increased from approximately 11% to 23% over a 23-year period, probably reflecting both a decrease in squamous carcinomas and an increase in adenocarcinomas.1 Cervical cancer screening programs have been effective in decreasing both the incidence and mortality from squamous carcinoma, but although adenocarcinoma mortality rates have declined, incidence rates have not.2 This may reflect the difficulty in detecting adenocarcinoma in situ, but also suggests that invasive adenocarcinoma is being detected at an earlier stage. Most cases of invasive adenocarcinoma evolve from

Table 4. Preconization Abnormalities and Method of Subsequent Cone Biopsy Initial treatment modality Preconization abnormality

Cold knife cone

LEEP excision

49 (86) 15 (83) 24 (92) 10 (77) 32 (74) 8 (62) 12 (92) 12 (71) 4 (22)

8 (14) 3 (17) 2 (8) 3 (23) 11 (26) 5 (38) 1 (8) 5 (29) 14 (78)

†

Glandular,* n (%) Papanicolaou only Papanicolaou ⫹ histology Histology only Mixed (glandular ⫹ squamous),*‡ n (%) Papanicolaou Papanicolaou ⫹ histology Histology only Squamous LEEP ⫽ loop electrical excisional procedure. * Subheadings refer to modality with which glandular lesion was detected. † P ⬍ .001, Fisher exact test, two-tailed, glandular compared with squamous. ‡ P ⬍ .001, Fisher exact test, two-tailed, mixed compared with squamous.

VOL. 100, NO. 2, AUGUST 2002

Shin et al

Preconization Findings of Adenocarcinoma In Situ

273

an adenocarcinoma in situ precursor located within or just proximal to the transformation zone.3,9 –11 There is a window of approximately 5 years between clinically detectable adenocarcinoma in situ and early invasive adenocarcinoma, indicative of an opportunity for screening and detection before invasion.3,12 However, the detection of adenocarcinoma in situ has been challenging for a number of reasons. Papanicolaou smears may be less sensitive than they are for squamous precursors because adenocarcinoma in situ may mimic endometrial cells or reactive endocervical cells.13 Also, benign conditions such as tubal metaplasia14,15 and cervical endometriosis16 may cytologically mimic adenocarcinoma in situ. Colposcopic evaluation and sampling are more difficult because of the location of adenocarcinoma in situ within the endocervical canal.10,17 However, ECC has been insensitive for detecting adenocarcinoma in situ.18 A coexisting SIL may obscure the glandular lesion, especially when extensive or high grade.17 Despite these challenges, our observation has been that the number of adenocarcinoma in situ cases that we have identified has increased with time. Our increasing rate of detection undoubtedly reflects the rising incidence of adenocarcinoma in situ, but also suggests improved recognition. Historically, only sporadic cases of adenocarcinoma in situ were reported in the two and a half decades after its first identification19 and description in 1953.20 In the 1970s and 1980s, a number of descriptive studies detailing the cytologic features of adenocarcinoma in situ were published,7,8,21,22 increasing the awareness and the diagnostic skill of cytopathologists. In the 1980s, introduction of new endocervical sampling brushes improved the ability to collect cells, both normal and abnormal, from the endocervical canal.23–25 These factors coincided with an improved rate of adenocarcinoma in situ detection, as illustrated by two case series that report 88 cases between 1980 and 1996. In these, only 12 (14%) cases were identified before 1987 compared with 76 (86%) after 1988.18,26 In a number of retrospective case series published within the last 10 years, screening Papanicolaou smears detected a glandular abnormality before confirmation of adenocarcinoma in situ on cone biopsy or hysterectomy in 32–55% cases.18,26 –29 The glandular abnormalities reported on cytology in these studies included adenocarcinoma in situ in 30% of cases, adenocarcinoma in 18%, atypical glandular cells in 29%, and glandular atypia in 23%. Follow-up evaluation with colposcopically directed biopsies in these retrospective studies revealed glandular abnormalities in 35–70% of proven adenocarcinoma in situ cases.18,26 –29 Also, ECC found a glandular abnormality in 17–35% of cases.18,26,28 In our review of 118 women with adenocarcinoma in

274

Shin et al

Preconization Findings of Adenocarcinoma In Situ

situ on a cone biopsy, the Papanicolaou smear had a sensitivity of 69% in detecting a glandular abnormality preoperatively. Thirty-one percent of glandular Papanicolaou smear findings were reported as adenocarcinoma in situ, 5% as adenocarcinoma, 55% as atypical glandular cells (61% suggesting adenocarcinoma in situ or adenocarcinoma), and 9% as glandular atypia. Combination of all cytologic and histologic testing in our study increased sensitivity to 85% of cases in which a glandular abnormality was anticipated at the time of initial treatment. Adenocarcinoma in situ was an incidental finding at the time of conization done for squamous disease in only 15% of cases. Our study has demonstrated that the ability to detect adenocarcinoma in situ with routine screening methods has continued to improve. The high frequency of preconization detection of glandular abnormalities observed in our patient population may be falsely elevated because of the high proportion of referred cases at Brigham and Women’s Hospital. Moreover, our current study design and that of others18,26 –29 does not permit an assessment of the positive predictive value of cytologic glandular abnormalities in detecting adenocarcinoma in situ. Nevertheless, our findings indicate that the majority of clinically detected adenocarcinoma in situ cases are preceded by either a cytologic or histologic interpretation of a glandular abnormality and that these are more effectively detected when present as a pure adenocarcinoma in situ lesion rather than mixed with SIL. Given the demonstrated high sensitivity of our screening tests in preoperatively detecting adenocarcinoma in situ, we would recommend a thorough evaluation of any patient with a suspected glandular lesion. The high correlation between preoperative detection of glandular abnormalities and the finding of adenocarcinoma in situ on cone biopsy has important implications for treatment choice. In our study, the majority of patients who had a glandular lesion suspected on cytology or histology underwent a cold knife cone biopsy as their initial treatment modality. When compared with loop electrical excision procedure, cold knife cone results in a larger and deeper specimen, and interpretation is not limited by cautery artifact,30 both important considerations in the management of adenocarcinoma in situ.26,31 We recently published data suggesting the safety of conservative management of adenocarcinoma in situ in women desiring future fertility if cone biopsy margins are negative.31 The sensitivities of our screening tests (Papanicolaou smear, biopsy, ECC) are critical to our ability to offer this option, both from the standpoint of preoperative planning as well as with postoperative surveillance. Clinicians at our institution had a bias towards treating suspected glandular lesions with cold knife cone

OBSTETRICS & GYNECOLOGY

rather than with loop excision. The success of conservative management in this patient population may correlate with the treatment decisions executed based upon the preoperative suspicion of glandular lesions. The new Bethesda system 2001 Papanicolaou smear terminology recommended several changes to the way glandular lesions are reported. “Atypical glandular cells of undetermined significance” will now be reported as “atypical glandular cells.” Additionally, “adenocarcinoma in situ” was created as a distinct category, and the category of “atypical glandular cells of undetermined significance-favor reactive” was eliminated in the new system. The intent in making these changes was to put more emphasis on glandular lesions as high-risk markers for significant pathology. The use of liquid-based Papanicolaou smears may increase the diagnostic specificity of glandular abnormalities for high-grade lesions.32 As our understanding of glandular lesions continues to expand and cervical sampling techniques continue to improve, we may expect continued enhancement in our ability to detect and treat adenocarcinoma in situ before its development into invasive adenocarcinoma. REFERENCES 1. Plaxe SC, Saltzstein SL. Estimation of the duration of the preclinical phase of cervical adenocarcinoma suggests that there is ample opportunity for screening. Gynecol Oncol 1999;75:55– 61. 2. Nieminen P, Kallio M, Hakama M. The effect of mass screening on incidence and mortality of squamous and adenocarcinoma of cervix uteri. Obstet Gynecol 1995;85: 1017–21. 3. Lee KR, Flynn CE. Early invasive adenocarcinoma of the cervix: A histopathologic analysis of 40 cases with observations concerning histogenesis. Cancer 2000;89:1048 –55. 4. Barron BA, Richart RM. Screening protocols for cervical neoplastic disease. Gynecol Oncol 1981;12(Suppl): S156 –S167. 5. Urcuyo R, Rome RM, Nelson JH. Some observations on the value of endocervical curettage performed as an integral part of colposcopic examination of patients with abnormal cervical cytology. Obstet Gynecol 1977;128: 787–92. 6. Jaworski RC, Pacey NF, Greenberg ML, Osborn RA. The histologic diagnosis of adenocarcinoma in situ and related lesions of the cervix uteri. Adenocarcinoma in situ. Cancer 1988;61:1171– 81. 7. Bousfield L, Pacey F, Young Q, Krumins I, Osborn RA. Expanded cytologic criteria for the diagnosis of adenocarcinoma in situ of the cervix and related lesions. Acta Cytol 1980;24:283–96. 8. Ayer B, Pacey F, Greenberg ML, Bousfield L. The cytologic diagnosis of adenocarcinoma in situ of the cervix

VOL. 100, NO. 2, AUGUST 2002

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20. 21.

22.

23.

24.

25.

uteri and related lesions: I. Adenocarcinoma in situ. Acta Cytol 1987;31:397– 411. Teshima S, Shimosato Y, Kishi K, Kasamatsu T, Ohmi K, Uei Y. Early stage adenocarcinoma of the uterine cervix: Histopathologic analysis with consideration of histogenesis. Cancer 1985;56:167–72. Ostor AG, Pagano R, Davoren RAM, Fortune DW, Chanen W, Rome R. Adenocarcinoma in situ of the cervix. Int J Gynecol Pathol 1984;3:179 –90. Noda K, Kimura K, Ikeda M, Teshima K. Studies on the histogenesis of cervical adenocarcinoma. Int J Gynecol Pathol 1983;1:336 – 46. Boon ME, Baak JPA, Kurver PJH, Overdiep SH, Verdonk GW. Adenocarcinoma in situ of the cervix: An underdiagnosed lesion. Cancer 1981;48:768 –73. Lee KR, Minter LJ, Granter SR. Papanicolaou smear sensitivity for adenocarcinoma in situ of the cervix: A study of 34 cases. Am J Clin Pathol 1997;107:30 –5. Novotny DB, Maygarden SJ, Johnson DE, Frable WJ. Tubal metaplasia: A frequent potential pitfall in the cytologic diagnosis of endocervical glandular dysplasia on cervical smears. Acta Cytol 1992;36:1–10. Wells M, Brown LJR. Glandular lesions of the uterine cervix: The present state of our knowledge. Histopathology 1986;10:777–92. Pacey F, Ayer B, Greenberg ML. The cytologic diagnosis of adenocarcinoma in situ of the cervix uteri and related lesions: III. Pitfalls in diagnosis. Acta Cytol 1988;32: 325–30. Christopherson WM, Nealon N, Gray LA. Noninvasive precursor lesions of adenocarcinoma and mixed adenosquamous carcinoma of the cervix uteri. Cancer 1979;44: 975– 83. Denehy TR, Gregori CA, Breen JL. Endocervical curettage, cone margins, and residual adenocarcinoma in situ of the cervix. Obstet Gynecol 1997;90:1– 6. Hepler TK, Dockerty MB, Randall LM. Primary adenocarcinoma of the cervix. Am J Obstet Gynecol 1952;63: 800 – 8. Friedell GH, McKay DG. Adenocarcinoma in situ of the endocervix. Cancer 1953;6:887–97. Krumins I, Young Q, Pacey F, Bousfield L, Mulhearn L. The cytologic diagnosis of adenocarcinoma in situ of the cervix uteri. Acta Cytol 1977;21:320 –9. Betsill WL Jr, Clark AH. Early endocervical glandular neoplasia: I. Histomorphology and cytomorphology. Acta Cytol 1986;30:115–26. Kristensen GB, Holund B, Grinsted P. Efficacy of the cytobrush versus the cotton swab in the collection of endocervical cells. Acta Cytol 1989;33:849 –51. Neinstein LS, Rabinovitz S, Recalde A. Comparison of cytobrush with cotton swab for endocervical cytologic sampling. J Adolesc Health Care 1989;10:305–7. Boon ME, de Graaft Guilloud JC, Rietveld WJ. Analysis

Shin et al

Preconization Findings of Adenocarcinoma In Situ

275

26.

27.

28.

29.

30.

276

of five sampling methods for the preparation of cervical smears. Acta Cytol 1989;33:843– 8. Widrich T, Kennedy AW, Myers TM, Hart WR, Wirth S. Adenocarcinoma in situ of the uterine cervix: Management and outcome. Gynecol Oncol 1996;61:304 – 8. Muntz HG, Bell DA, Lage JM, Goff BA, Feldman S, Rice LW. Adenocarcinoma in situ of the uterine cervix. Obstet Gynecol 1992;80:935–9. Wolf JK, Levenback C, Malpica A, Morris M, Burke T, Mitchell MF. Adenocarcinoma in situ of the cervix: Significance of cone biopsy margins. Obstet Gynecol 1996;88: 82– 6. Azodi M, Chambers SK, Rutherford TJ, Kohorn EI, Schwartz PE, Chambers JT. Adenocarcinoma in situ of the cervix: Management and outcome. Gynecol Oncol 1999; 73:348 –53. Mathevet P, Dargent D, Roy M, Beau G. A randomized prospective study comparing three techniques of coniza-

Shin et al

Preconization Findings of Adenocarcinoma In Situ

tion: Cold knife, laser, and LEEP. Gynecol Oncol 1994; 54:175–9. 31. Shin CH, Schorge JO, Lee KR, Sheets EE. Conservative management of adenocarcinoma in situ of the cervix. Gynecol Oncol 2000;79:6 –10. 32. Hecht JL, Sheets EE, Lee KR. Atypical glandular cells of undetermined significance (AGUS) in conventional cervical/vaginal smears and thin-layer preparations. A follow-up comparison study. Cancer Cytopathol 2002;96: 1– 4. Address reprint requests to: Ellen E. Sheets, MD, Brigham and Women’s Hospital, Division of Gynecologic Oncology, 75 Francis Street, Boston, MA 02115; E-mail: esheets@partners. org. Received November 8, 2001. Received in revised form February 14, 2002. Accepted March 7, 2002.

OBSTETRICS & GYNECOLOGY