Calcifications in Ovary and Endometrium and Their Neoplasms Elvio G
Total Page:16
File Type:pdf, Size:1020Kb
Calcifications in Ovary and Endometrium and Their Neoplasms Elvio G. Silva, M.D., Michael T. Deavers, M.D., A. F. Parlow, M.D., David M. Gershenson, M.D., Anais Malpica, M.D. Departments of Pathology (EGS, AM, MTD) and Gynecologic Oncology (DMG), The University of Texas M. D. Anderson Cancer Center, Houston, Texas; and Department of Pathology (AFP), Harbor-UCLA Medical Center, Torrance, California The third part of the study was directed to deter- In this study, we investigated the role of hormones mine the frequency of calcifications in ovarian le- in the pathogenesis of calcifications in ovary and in sions. We found that all cases of endosalpingiosis endometrium and their neoplasms of the gyneco- and ovarian low-grade serous carcinoma had calci- logic tract and assessed the anatomic location and fications, whereas 80% of the cases of serous bor- incidence of these calcifications. The study consists derline tumor had calcifications, and only 50% of of three parts designed to investigate the pathogen- the cases of ovarian high-grade serous carcinoma esis, the location, and the incidence of calcifications contained calcifications. The results of this study in ovary and endometrium and their neoplasms. In indicate that the majority of the calcifications in the the first part, 79 female guinea pigs were divided ovary and the endometrium and their neoplasms into 10 groups, and different hormones, given are present in the stroma. This is most probably weekly for 12 months, were administered to the secondary to metabolic changes, which could be guinea pigs by group. A control group of 7 guinea related to hormones and not caused by degenera- pigs received sterile water. Calcifications developed tive changes in epithelial cells. in 5 of 7 guinea pigs treated with prolactin, 10 of 20 treated with human chorionic gonadotropin, 5 of 11 KEY WORDS: Calcifications, Endometrium, Hor- treated with estradiol, 3 of 7 treated with estrone, 1 mones, Ovary, Pathogenesis, Tumor. of 6 treated with growth hormone, and 1 of 10 Mod Pathol 2003;16(3):219–222 treated with testosterone; in 20 of the guinea pigs, the calcifications developed in the stroma of the Pathologic calcifications are classified as either endometrium, and in 5 guinea pigs, they developed metastatic (associated with hypercalcemia) or dys- in the ovary. The second part of the study consisted trophic (associated with normal calcemia) (1, 2). of an evaluation of the specific location of calcifica- Traditionally, calcifications in neoplasms have been tions in 43 consecutive human surgical ovaries and considered to be dystrophic, forming secondary to endometria. Calcifications were seen only in the degeneration of the epithelium or in association stroma in 100% of the ovarian serous adenofibroma with areas of necrosis (2–6). However, a secretory specimens; in ovarian serous borderline neoplasms, rather than a degenerative etiology has also been the stroma contained 70 to 100% of the calcifica- suggested on the basis of the absence of stainable tions, and the epithelium had 0 to 30% of the calci- DNA and RNA in areas with calcifications (7–9). fications. In ovarian serous carcinoma specimens, In this study, we investigated the role of hor- the calcifications were seen in the stroma in 50 to mones in the pathogenesis of calcifications in the 60% of the cases, in the epithelium in 40% of the ovary and the endometrium and their neoplasms cases, and in areas of necrosis in 10% of the cases. and assessed the anatomic site and incidence of these calcifications. A novel pathogenetic mecha- Copyright © 2003 by The United States and Canadian Academy of nism to explain the presence of calcifications in the Pathology, Inc. ovary and endometrium is presented. VOL. 16, NO. 3, P. 219, 2003 Printed in the U.S.A. Date of acceptance: January 8, 2003. Supported in part by the Polo on the Prairie Grant Foundation and the Ovarian Cancer Research Program. MATERIALS AND METHODS Address reprint requests to: Elvio G. Silva, M.D., Department of Pathology, Box 85, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030; fax: 713-792-5529; e-mail: This study was divided into three parts to study [email protected]. the pathogenesis, the location, and the incidence of DOI: 10.1097/01.MP.0000057236.96797.07 calcifications in the ovary and the endometrium 219 and their neoplasms. To study the pathogenesis of in a retrospective review of 56 serous lesions, in- calcifications, 79 female guinea pigs were divided cluding 10 cases of ovarian endosalpingiosis and 46 into 10 different groups, and a different hormone cases of ovarian neoplasms (10 serous borderline was administered to each group. The hormones, tumors, 16 low-grade serous carcinomas, and 20 the weekly doses and modes of administration, and high-grade serous carcinomas). the number of guinea pigs per group were as fol- The low- and high-grade serous carcinomas were lows, respectively: estradiol, 1 mg intramuscularly classified by criteria recently proposed by Malpica (IM), n ϭ 11; estrone, 1.2 mg IM, n ϭ 7; testoster- et al. (12). Serous carcinomas with Յ10 mitoses per one, 80 mg IM, n ϭ 10; dihydrotestosterone, 10 mg 10 high-power fields or absence of significant nu- orally, n ϭ 4; tamoxifen, 3 mg orally, n ϭ 5; meges- clear pleomorphism were classified as low-grade trol acetate, 100 mg orally, n ϭ 4; clomiphene, 1.5 tumors, and those with Ͼ10 mitoses per 10 high- mg orally, n ϭ 5; human chorionic gonadotropin, power fields or with significant nuclear pleomor- 500UIM,n ϭ 20; prolactin, 0.4 mL IM, n ϭ 7; and phism were classified as high-grade tumors. growth hormone, 0.2 mL IM, n ϭ 6. All of the hormones were administered for 12 months. The number of animals placed in each group was de- RESULTS termined by our previous experience with these hormones (10, 11). Larger numbers of animals were Calcifications were found in 25 of the 79 guinea placed in the groups receiving hormones likely to pigs that received hormones. No calcifications were induce calcifications, whereas a smaller number of found in the guinea pigs that received sterile water. animals received the hormones that were less likely The number of guinea pigs with calcifications, by to induce calcifications, or hormones that were dif- the hormone received, was as follows: estradiol, 5 of ficult to obtain, such as prolactin and growth hor- 11 guinea pigs; estrone, 3 of 7 guinea pigs; testos- mone. The control group comprised 7 guinea pigs terone, 1 of 10 guinea pigs; dihydrotestosterone, 0 that received sterile water in doses of 2 mL/wk IM of 4 guinea pigs; tamoxifen, 0 of 5 guinea pigs; for 12 months. After 12 months, all of the guinea megestrol, 0 of 4 guinea pigs; clomiphene, 0 of 5 pigs were killed, and a complete autopsy was per- guinea pigs; human chorionic gonadotropin, 10 of formed on each. 20 guinea pigs; prolactin, 5 of 7 guinea pigs; and To evaluate the locations of calcifications, we growth hormone, 1 of 6 guinea pigs. The calcifica- reviewed 43 consecutive human gynecologic spec- tions were found in the stroma of the endometrium imens: 33 epithelial ovarian neoplasms, including 3 and within endometrial glands in 20 of the 25 cases, serous adenofibromas, 6 serous borderline tumors, and in the ovarian stroma in the remaining 5 cases. 13 low-grade serous carcinomas, 10 high-grade se- Each of these 25 cases had one to five foci contain- rous carcinomas, and 1 undifferentiated carcino- ing calcifications and had no areas of necrosis as- ma; and 10 endometria, including 5 proliferative sociated with these calcifications. endometrium, 3 endometrial hyperplasia, and 2 en- In the 43 gynecologic specimens reviewed to de- dometrial endometrioid adenocarcinomas with cal- termine the location of the calcifications, the sites cifications. We evaluated calcifications smaller than varied by the histologic type of lesion. Table 1 30 m in greatest dimension because the precise shows the results of the 33 ovarian neoplasms. In origin of calcifications larger than 30 m generally the three ovarian serous adenofibromas, the calci- could not accurately be determined. In specimens fications were present only in the stroma. Necrosis with small calcifications, we evaluated the location and mitoses were not seen in these areas. In the six of the calcifications, the presence of necrosis or ovarian serous borderline tumors, the percentage of vascular changes, and the number of mitoses in the calcifications in each tumor present in the epithe- cells near the calcifications. lium ranged from 0 to 30% (median, 0%); con- To assess the frequency of calcifications, we in- versely, the percentage in the stroma ranged from vestigated the presence or absence of calcifications 100 to 70% (median, 100%; Fig. 1). No necrosis or TABLE 1. Histologic Features, Distribution, and Frequency of Calcifications in 33 Human Epithelial Ovarian Neoplasms Percentage of Calcifications by Site, Range (Median) Number of Histologic Diagnosis Number of Cases (n) Area of Mitoses per 10 Epithelium Stroma Necrosis Necrosis HPF, Median Serous adenofibroma, n ϭ 3 0 (0) 100 (100) —— 0 Serous borderline, n ϭ 60–30 (0) 70–100 (100) —— 0 Low-grade serous carcinoma, n ϭ 13 10–60 (40) 40–80 (60) 5–10 (7) ϩ 3 High-grade serous carcinoma, n ϭ 10, 20–70 (37) 30–80 (50) 4–40 (10) ϩϩϩ 21 and undifferentiated carcinoma, n ϭ 1 Necrosis: ϩϭone focus Ͻ 3 mm; ϩϩ ϭ two to three foci, Ͻ 3 mm; ϩϩϩ ϭ over three foci or larger than 3 mm. 220 Modern Pathology FIGURE 1. Ovarian serous borderline tumor. Small calcifications in FIGURE 3. Serous carcinoma in omentum. Calcifications in the the stroma; several large calcifications at the center of the papillae.