Endometrial Adenocarcinoma in Situ in Complex Atypical Hyperplasia: Correlation with Findings in Subsequent Hysterectomy Specimen

Transcription

1 International Journal of Surgical Pathology 12(3): , 2004 Endometrial Adenocarcinoma in Situ in Complex Atypical Hyperplasia: Correlation with Findings in Subsequent Hysterectomy Specimen Karyna C. Ventura, MD, Dorota Popiolek, MD, and Khush Mittal, MD Well-differentiated endometrial adenocarcinoma can be difficult to distinguish from complex atypical hyperplasia (CAH) in a curettage or biopsy specimen. When a focus of back-to-back glands or cribriforming smaller than 2.1 mm is seen in a biopsy, we make a diagnosis of adenocarcinoma in situ (AIS). Whether this diagnosis translates into a more frequent diagnosis of carcinoma on the hysterectomy specimen is unknown. The objective of this study was to compare follow-up hysterectomy findings in biopsies showing AIS in CAH with biopsies showing only CAH without AIS. Twelve biopsy/curettage cases diagnosed as endometrial AIS in CAH and 12 biopsy/curettage cases diagnosed as CAH only were reviewed and correlated with corresponding hysterectomy material. A diagnosis of AIS was designated on biopsy/curettings when a focus of back-to-back glands or cribriforming less than 2.1 mm was present. Hysterectomy specimens showed endometrial carcinoma in 6 (50%) of 12 cases of CAH with AIS, and in 2 (17%) of 12 cases diagnosed as CAH only. Endometrial carcinoma with myometrial invasion was identified in 5 (42%) of the cases showing AIS on biopsy, but in none of the 12 cases diagnosed as CAH only on biopsy. Identification of AIS in CAH cases provides useful prognostic information. Int J Surg Pathol 12(3): , 2004 Key words: adenocarcinoma in situ, complex endometrial hyperplasia with atypia, endometrial carcinoma. Well-differentiated adenocarcinoma of the endometrium (EM) can be difficult to distinguish from complex atypical endometrial hyperplasia. In the spectrum of endometrial proliferation, endometrial adenocarcinoma in situ (AIS) is a preinvasive endometrial disease that presumably involves neoplastic endometrial glands without identifiable invasion of endometrial stroma [1 4]. Its diagnostic New York University School of Medicine, New York, NY. Reprint requests: Khush Mittal, MD, Director of Obstetric-Gynecologic Pathology, Room 4W35, Department of Pathology, NYU School of Medicine/Bellevue Hospital Center, First Avenue at 27th Street, New York, NY criteria are not clearly defined and its existence and recognition remain in question [1,5,6]. Some authors have equated AIS with complex endometrial hyperplasia with atypia, whereas others prefer terms such as intraendometrial carcinoma, adenocarcinoma without stromal invasion, marked adenomatous hyperplasia, or intraendometrial neoplasia to describe this lesion [1,2,7]. Kurman and Norris [8] recommend that endometrial adenocarcinoma be diagnosed only when the following processes involve at least half of a low-power field (or 2.1 mm), indicating stromal invasion: (1) irregular infiltration of glands into desmoplastic stroma, (2) a confluent glandular/cribriform pattern, (3) an extensive papillary pattern, and (4) replacement of the stroma by 225

2 226 International Journal of Surgical Pathology Vol. 12 No. 3 July 2004 masses of squamous epithelium. Complex endometrial hyperplasia with atypia, alternatively, represents an endometrial proliferation morphologically identified by complex or markedly crowded glands without evidence of stromal invasion as described above [9]. Well-differentiated adenocarcinoma can be especially difficult to distinguish from complex atypical endometrial hyperplasia in a curettage or biopsy specimen. When back-to-back glands or cribriforming are seen in a focus smaller than 2.1 mm, we make a diagnosis of adenocarcinoma in situ (AIS). Whether this diagnosis translates into a worse finding on the hysterectomy specimen is unknown. The objective of our study is to compare followup hysterectomy findings in cases where biopsies show AIS in complex atypical hyperplasia (CAH) with hysterectomy findings in cases where biopsies show only CAH without AIS. Materials and Methods Endometrial biopsy/curettage cases with diagnosis of endometrial adenocarcinoma in situ (AIS) in a background of complex atypical hyperplasia (CAH), dating from , were retrieved from the archives of the New York University Medical Center and Bellevue Hospital Centers. Only cases treated with subsequent hysterectomy were included in this study. These biopsy/curettage cases were reviewed and correlated with corresponding hysterectomy findings. Adenocarcinoma in situ was identified by the presence of back-to-back arrangement of glands or cribriforming that did not exceed 2.1 mm and that was composed of at least 4 glands (Figs. 1 3). Cases meeting criteria for stromal invasive endometrial adenocarcinoma were excluded. Stromal invasion is defined by the presence of the following features in a focus greater than 2.1 mm: (1) irregular infiltration of glands into desmoplastic stroma, (2) a confluent glandular pattern/cribriform, (3) an extensive papillary pattern, and (4) replacement of the stroma by masses of squamous epithelium [8]. Adenocarcinoma in situ cases were compared with another group of biopsy/curettage cases diagnosed with only complex atypical endometrial hyperplasia. All endometrial biopsy/curettage slides, as well as hysterectomy slides, had been previously processed and stained routinely with hematoxylin and eosin 1A 2A 3A 1B 2B 3B Figs. 1 to 3. Three examples of complex atypical hyperplasia with adenocarcinoma in situ. The top part of each figure (A) is lowpower view to illustrate the background of complex hyperplasia. The bottom part of each figure (B) shows high-power appearance of foci of adenocarcinoma in situ. (Hematoxylin and Eosin, 40 and 200 ).

3 Endometrial AIS Ventura et al. 227 stain. Endometrium was submitted in toto from all hysterectomy specimens. All biopsy/curettage materials were retrospectively reviewed. The following features were evaluated: size of the largest AIS focus, number of glands in the largest AIS focus, number of AIS foci, mitotic index, nuclear size, presence of nucleoli, and nucleus-to-cytoplasmic ratio. An additional control group of 15 randomly selected cases with diagnosis of FIGO grade I endometrial adenocarcinoma on endometrial curettings, and on whom follow-up hysterectomy results were available, were evaluated for frequency of myoinvasive adenocarcinoma. Statistical analysis was performed using Fisher s Exact Test. Results A total of 12 endometrial biopsy/curettage cases with AIS were reviewed and compared with a group of 12 endometrial biopsy/curettage cases with diagnosis of CAH. The patients in the AIS group had a mean age of years (SD ± 7.3) while the patients in the CAH group had a mean age of 53.5 years (SD ± ). There is no significant difference between the mean ages of the 2 groups. Six of the 12 (50%) AIS cases contained endometrial adenocarcinoma (ACA) in their corresponding hysterectomy material. In contrast, only 2 of the 12 cases of CAH without AIS (17%) contained endometrial adenocarcinoma in their hysterectomy specimens. The difference did not, however, reach statistical significance (p = 0.19, Table 1). There was, however, a statistically significantly higher myometrial invasion in the AIS group. Five of the 12 cases from the AIS group showed myometrial invasion (13%, 14%, 25%, 42%, and 70%, respectively, of myometrial wall thickness); while none of the 12 cases from the CAH group demonstrated myometrial invasion (p = 0.03, Table 1). Three of the CAH cases and 1 of the AIS cases demonstrated only adenocarcinoma in situ in their corresponding hysterectomy specimens. One case from AIS group contained only normal proliferative endometrium on the hysterectomy specimen. On evaluation of the microscopic features of these endometrial biopsies/curettages, the AIS cases contained 0 to 28 mitotic figures per 10 high-power field (hpf) with an average of 6 mitoses/10 hpf. The CAH group had mitotic counts that ranged from 0 to 15/10 hpf with an average of 4 mitoses/10 hpf. There was no significant association between the number of cells in mitosis and the presence of endometrial adenocarcinoma with myometrial invasion (Table 2). Additionally, nuclear size, nucleoli, and nucleusto-cytoplasmic ratio were not predictive of the presence of myoinvasive endometrial carcinoma (Table 2). The foci of AIS in the biopsies/curettage cases studied ranged from 0.3 to 1 mm with 4 to 26 glands per focus. The size of the AIS focus/foci, the number of AIS foci, or the number of glands in the AIS foci could not significantly predict the presence of myoinvasive endometrial carcinoma in subsequent hysterectomy specimens (Table 2). In the control group of 15 patients with diagnosis of FIGO grade I endometrial adenocarcinoma on endometrial curettage, myoinvasive adenocarcinoma was present in 13 cases (Table 1). This was a significantly greater incidence of myoinvasive adenocarcinoma than that seen with the AIS group (5/12 vs. 13/15, p = 0.02). One patient in this group had carcinoma confined to the endometrium, while another showed no residual carcinoma. Discussion The recognition and existence of endometrial adenocarcinoma in situ has been controversial [1,5,6]. In the past, AIS has been used to describe a variety of lesions that are now recognized as Table 1. Findings in Endometrial Biopsies and Corresponding Hysterectomy Specimens Hysterectomy Findings Biopsy Results Normal CAH AIS ACA: Confined to EM ACA: Myoinvasive Total Complex atypical hyperplasia (CAH) Adenocarcinoma in situ (AIS) * 12 FIGO grade I endometrial adenocarcinoma *p value: 0.03 versus CAH, and 0.02 versus Grade I adenocarcinoma significant; ACA = adenocarcinoma; EM = endometrium.

4 228 International Journal of Surgical Pathology Vol. 12 No. 3 July 2004 Table 2. Cytologic Features in Preceding Biopsy of Patients with CAH with Myoinvasive Carcinoma on Subsequent Hysterectomy Myoinvasive ACA Present Absent p Value Prominent nucleoli: Present NS Absent 4 14 Mitosis: >10/10hpf NS <10/10hpf 4 16 Mitosis: >5/10hpf NS <5/10hpf 4 14 Nuclear size ( lymphocytes): >3x NS <3x 3 18 N/C ratio: >50% NS <50% 3 17 Size of AIS: >0.5 mm NS <0.5 mm 1 2 No. of glands in AIS: > NS < NS = not significant. eosinophilic metaplasia, epithelial intraepithelial neoplasia, and complex atypical endometrial hyperplasia [9]. Previously, the term marked adenomatous hyperplasia was also used for these lesions [2]. AIS has been described to have loss of polarity and piling up of epithelial cells, intraglandular bridges with nuclear hyperchromatism and irregularity, clumping of nuclear chromatin, and cytoplasmic eosinophilia [3,4]. Welch and Scully [10] described adenocarcinoma in situ as small focal lesions involving no more than 5 or 6 glands in which the features of carcinoma are present without evidence of invasion into stroma. No size criterion for the diagnosis of AIS had been put forward previously. If we were to apply the definition of Kurman and Norris [8] to the spectrum of endometrial glandular lesions, there would be foci of endometrial glandular proliferation that would morphologically appear as adenocarcinoma but would show no stromal invasion. King et al [7] described these cases as adenocarcinoma without stromal invasion. The diagnosis of AIS should be distinguished from endometrial intraepithelial carcinoma (EIC), which is characterized by a single layer of cells with highgrade nuclei, frequent p53 expression, and association with papillary serous carcinoma [11]. The diagnosis of AIS should also be distinguished from endometrial intraepithelial neoplasia (EIN) [12]. Owing to difficulty in reproducing diagnoses using the present WHO nomenclature for endometrial hyperplasia/carcinoma, Mutter and the Endometrial Collaboration Group [12] have advocated the use of endometrial intraepithelial neoplasia, or EIN terminology. The division of preinvasive endometrial carcinoma into endometrial hyperplasia and endometrial intraepithelial neoplasia, it is argued, supports a two-tiered approach to the treatment of endometrial lesions, with medical therapy recommended for endometrial hyperplasia and medical or surgical treatment for EIN [12,13]. Silverberg [5] points out, however, that this system will then herald the new dilemma of deciding when a hyperplasia is severe enough to qualify as EIN, which would indicate surgery, and more importantly, when a carcinoma becomes severe enough to be diagnosed beyond EIN. In keeping with the nomenclature of preinvasive lesions of the cervix, vagina, and vulva, AIS of the endometrium would represent a preinvasive lesion beyond EIN that would indicate a high risk for the presence or development of carcinoma and thus would require surgical treatment. AIS and Myometrial Invasion Among the different prognostic factors considered in the evaluation of endometrial carcinoma, myometrial invasion is considered to be the most clinically significant parameter [5,7]. Therefore, the identification of a feature or a lesion on biopsy ma-

5 Endometrial AIS Ventura et al. 229 terial that may prove to be associated with myometrial invasion must be fully evaluated. Our results demonstrate the significant association of AIS on endometrial biopsy with the presence of myoinvasive endometrial adenocarcinoma in subsequent hysterectomy specimens, in comparison to a diagnosis of only complex atypical hyperplasia on biopsy material. These findings suggest a possible prognostic and therapeutic significance of the diagnosis of AIS on biopsy material. King et al [7] showed that 2% (2/119) of their cases of atypical hyperplasia had myometrial invasive adenocarcinoma on hysterectomy, while 16% (7/43) of their cases of adenocarcinoma without stromal invasion (what we call AIS) on biopsy had myometrial invasion in the hysterectomy specimen (2% vs 16%). In their cases of invasive adenocarcinoma on the endometrial biopsies, myometrial invasion was seen in 62.5% of the cases [7]. In a study of 89 endometrial curettings with CAH without stromal invasion, Kurman and Norris [8] showed that only 17% of subsequent hysterectomies contained adenocarcinoma, with myometrial invasion in 8% of the cases. This group includes, however, cases of complex atypical hyperplasia as well as adenocarcinoma in situ diagnosed on biopsy. Findings in cases with CAH alone or CAH with adenocarcinoma without stromal invasion (what we call AIS) were not specified in that study. In their cases of invasive endometrial carcinoma by biopsy, adenocarcinoma was seen in 50% of the cases, with myometrial invasive carcinoma in 23% of the cases [14]. AIS with Subsequent CAH and Normal Endometrium Although one can argue that these small foci of adenocarcinoma are truly focal areas of well-differentiated adenocarcinoma, our study shows that some of the cases that we diagnosed as AIS on biopsy contained only complex atypical hyperplasia (4 cases) or even normal proliferative endometrium (1 case) in the hysterectomy specimen. King et al [7] found that 30% of their biopsy cases of adenocarcinoma without stromal invasion had no residual tumor on hysterectomy specimen. This may suggest the lesions were treated by the biopsy; or that the lesions were focal and not represented on the slides made from the hysterectomy specimen. Complex Atypical Hyperplasia Two of our 12 (17%) cases of CAH without AIS contained adenocarcinoma in the hysterectomy material. Our findings are similar to those of King et al [7], who found carcinoma in 15% of their CAH cases. Two of their 119 atypical hyperplasia cases showed adenocarcinoma with myometrial invasion, whereas none of our cases of CAH without AIS showed adenocarcinoma with myometrial invasion. Kurman and Norris [8] reported that 29% of their complex atypical hyperplasia cases contained well-differentiated adenocarcinoma in subsequent hysterectomy. Tavassoli and Kraus [6], similarly, reported that 25% of their atypical hyperplastic cases contained well-differentiated adenocarcinoma in subsequent hysterectomy specimens. Their cases included, however, both CAH and AIS cases, which may account for the higher percentage of adenocarcinoma compared to the King series and our series. More than 50% of our cases of CAH without AIS (7/12) had just CAH in the hysterectomy specimens while 3/12 cases showed AIS in the hysterectomy specimens (Table 1). CAH is currently treated with hysterectomy or progestational agents [15]. The findings of the current study suggest that patients with AIS in CAH should be treated by surgery, while those without AIS may by treated with progestational agents. AIS Versus FIGO Grade I Endometrial Adenocarcinoma The risk of myoinvasive carcinoma was significantly less in the AIS group than in the FIGO grade I endometrial adenocarcinoma group (5/12 versus 13/15, p = 0.02), indicating that AIS represents an intermediate risk category between CAH and FIGO grade I endometrial adenocarcinoma. In summary, our findings illustrate that a diagnosis of endometrial adenocarcinoma in situ on biopsy or curettage specimens is associated with endometrial adenocarcinoma in subsequent hysterectomy material more than half of the time. Additionally, a diagnosis of AIS on an endometrial biopsy significantly predicts the presence of myoinvasive endometrial adenocarcinoma, compared to a biopsy diagnosis of complex atypical endometrial hyperplasia without AIS. This suggests a possible prognostic significance of AIS in curettage or biopsy specimens. The number of cases in this study is small, and conclusions would need to be validated by additional studies. References 1. Fox H, Buckley CH. The endometrial hyperplasias and their relationship to endometrial neoplasia. Histopathology 6: , 1982

6 230 International Journal of Surgical Pathology Vol. 12 No. 3 July Gusberg SB, Kaplan AL. Precursors of corpus cancer. IV. Adenomatous hyperplasia as stage 0 carcinoma of the endometrium. Am J Obstet Gynecol 87: , Hertig AT, Sommers SC, Bengloff H. Genesis of endometrial carcinoma. III. Carcinoma in situ. Cancer 2: , Vellios F. Endometrial hyperplasia and carcinoma in situ. Gynecol Oncol 2: , Silverberg SG. Hyperplasia and carcinoma of the endometrium. Semin Diagn Pathol 5: , Tavassoli F, Kraus FT. Endometrial lesions in uteri resected for atypical endometrial hyperplasia. Am J Clin Pathol 70: , King A, Seraj IM, Wagner RJ. Stromal invasion in endometrial adenocarcinoma. Am J Obstet Gynecol 149:10 14, Kurman RJ, Norris H. Evaluation of criteria for distinguishing atypical endometrial hyperplasia from well-differentiated carcinoma. Cancer 49: , Abell MR. Adenocarcinoma (gland-cell carcinoma) in situ of endometrium. Pathol Res Pract 174: , Welch WR, Scully RE. Precancerous lesions of the endometrium. Hum Pathol 8: , Ambros RA, Sherman ME, Zahn CM, et al. Endometrial intraepithelial carcinoma: A distinctive lesion specifically associated with tumors displaying serous differentiation. Hum Pathol 26: , Mutter GL and The Endometrial Collaborative Group. Endometrial intraepithelial neoplasia (EIN): Will it bring order to chaos? Gynecol Oncol 76: , Ferenczy A, Gelfand M. Hyperplasia vs. neoplasia: Two tracks for the endometrium? Contemp Ob Gyn 28:79 96, Kurman RJ, Kaminsi PF, Norris HJ. The behavior of endometrial hyperplasia: A long-term study of untreated hyperplasia in 170 patients. Cancer 56: , Jobo T, Kawaguchi M, Imai M, Kuramoto H. Treatment for complex atypical hyperplasia of the endometrium. Eur J Gynaecol Oncol 22: , 2001