International Journal of Gynecological Pathology 33:23–29, Lippincott Williams & Wilkins, Baltimore r 2013 International Society of Gynecological Pathologists
Case Report
Involuting Luteinized Thecoma of the Ovary Lawrence M. Roth, M.D., Arthur R. Gaba, M.D., Michael D. Linden, and George E. Sandusky, D.V.M., Ph.D.
M.D.,
Summary: Spontaneous tumor involution in ovarian stromal tumors is a poorly understood phenomenon. In this report, we describe a rare case of luteinized thecoma that showed extensive involutional changes, such that cellular elements diagnostic of luteinized thecoma were sparse. The convoluted contour of the tumor resembled that observed in a corpus albicans; however, the neoplasm was considerably larger, and the contents of the nodule differed from that of a corpus albicans. The diagnosis of luteinized thecoma was established by the identification of residual aggregates of neoplastic theca cells and a nodule of lutein cells that were positive for inhibin and steroidogenic factor-1. Features of involution within the tumor included a few theca and lutein cells with pyknotic nuclei and abundant cytoplasmic lipid, occasional large adipocytes among the lutein cells, extensive hyalinization, dystrophic calcification, a myxohyaline nodule, and adipose metaplasia. It is likely that some of the aforementioned changes are the result of accompanying ischemia. Cleaved caspase-3 staining patterns were negative within residual lutein and theca cells; thus, we were unable to establish the occurrence of apoptotic bodies. Key Words: Ovary—Luteinized thecoma—Involution.
histologic changes that occur during the conversion of the corpus luteum to a corpus albicans and to those untreated unilateral acoustic schwannomas that decreased in size when followed for extended time periods by imaging techniques (3,4).
Luteinized thecoma is an uncommon, usually benign ovarian stromal tumor. In 1982, Zhang et al. (1) reported the largest series of 46 cases, 4 of which were malignant. The following year, 4 additional benign cases were described (2). Luteinized thecomas occur in younger women more often compared with typical thecomas, and 30% arise in women below 30 years of age (1). In this article, we use the term involution to describe secondary changes that occurred in the neoplasm we are reporting. This term has been applied previously to describe the
MATERIALS AND METHODS The tissue blocks were fixed in formalin, routinely processed, and embedded in paraffin. The slides were deparaffinized twice in xylene for 5 minutes, rehydrated using graded ethanol solutions to distilled water, and stained with hematoxylin and eosin. All slides were retrospectively reviewed and diagnosed according to the established criteria. Four-mm-thick sections were cut for immunohistochemical staining that was performed on a Dako Autostainer Plus (Dako, Carpinteria, CA). In brief, endogenous peroxidase activity was blocked using 3% hydrogen peroxide in methanol, and indigenous biotin activity was blocked using a protein blocker.
From the Departments of Pathology, Indiana University School of Medicine (L.M.R., G.E.S.), Indianapolis, Indiana; Henry Ford Health System (A.R.G.), Detroit, Michigan; and University of California San Diego School of Medicine and Veterans Administration San Diego Healthcare System (M.D.L.), San Diego, California. The authors declare no conflict of interest. Address correspondence and reprint requests to Lawrence M. Roth, MD, Department of Pathology, Indiana University School of Medicine, Van Nuys Medical Science Building 128, 635 Barnhill Drive, Indianapolis, IN 46202-5120. E-mail: [email protected].
23
DOI: 10.1097/PGP.0b013e31827db06f
24
L.M. ROTH ET AL.
After antigen retrieval, the slides were incubated with primary antibody. The detection systems used were Dako EnVision+, EnVision FLEX, and EnVision FLEX+. The peroxidase activity was developed with 3,3-diaminobenzidine and counterstained with hematoxylin. Appropriate positive controls for each antibody were run concurrently and showed adequate immunostaining. Slides of the neoplasm and an adjacent corpus luteum were analyzed for immunoreactivity using the following antibodies: inhibin a (inhibin), steroidogenic factor-1 (SF-1), calretinin, a-smooth muscle actin (SMA), cleaved caspase-3, Ki-67, and B-cell lymphoma 2 oncoprotein (BCL-2). Table 1 indicates the types of antibody used, the clone used for monoclonal antibodies, dilution, source, pretreatment, and the detection method. CASE REPORT A 49-year-old woman presented with a history of heavy, painful periods. She was G2 P1001. Her last menstrual period occurred 1 week before presentation. Transabdominal and endovaginal imaging of the pelvis showed a retroflexed uterus that measured 9 3.4 4.7 cm. The endometrial stripe measured 6.8 mm. The uterus and ovaries were unremarkable. The impression was an unremarkable pelvic ultrasound. At the patient’s request, a robotic-assisted hysterectomy and a bilateral salpingo-oophorectomy were performed for the relief of symptoms. There was no evidence of disease 1 month after the surgery. The specimen obtained from the surgery included the right ovary that measured 3.5 2.5 2.0 cm and an attached fallopian tube that measured 6 cm in
length. The external surface of the ovary was yellowtan in color and cerebriform in appearance. The sectioned surface showed a well-circumscribed tanwhite and yellow nodule measuring 1.7 1.4 1.3 cm. In addition, a hemorrhagic, cystic corpus luteum was identified that measured 1.5 1.5 1.1 cm. The uterus weighed 115.5 g, measured 8.0 6.5 4.2 cm, and showed no abnormalities. The left ovary measured 2.5 2.0 1.0 cm, and the fallopian tube measured 6.5 cm in length. None of these showed any abnormalities. Histologically, the right ovarian neoplasm was sharply demarcated from the surrounding ovarian stroma, and the surface of the neoplasm had a convoluted appearance that mimicked a corpus albicans (Fig. 1A). In some areas, aggregates of plump, spindle-shaped cells with elongated nuclei resembling thecoma were surrounded by hyalinized connective tissue (Figs. 1B, C). A nodule of lutein cells containing several large adipocytes was enveloped by hyalinized connective tissue (Fig. 1D). Adipose tissue was prominent and contained occasional clusters of residual theca cells (Fig. 1E). A thin band of hyaline connective tissue surrounded a single myxohyaline nodule (Fig. 1F). The tumor was sharply circumscribed and compressed the surrounding ovarian stroma (Fig. 2A). Patchy dystrophic calcification sometimes occurred within hyaline bands in association with involuting theca cells (Fig. 2B). Residual clusters of theca cells were surrounded by hyalinized connective tissue and adipocytes (Fig. 2C). Most theca cells had bland vesicular nuclei, but some had pyknotic nuclei and abundant vacuolated cytoplasm (Fig. 2D). The theca and lutein cells showed cytoplasmic staining for
TABLE 1. Details for antibodies and conditions used in this study Antibody Inhibin a SF-1 Calretinin a-Smooth muscle actin Cleaved caspase-3 Ki-67 BCL-2
Type Mouse monoclonal Mouse monoclonal Mouse monoclonal Mouse monoclonal Rabbit polyclonal Mouse monoclonal Mouse monoclonal
Clone R1 N1665
Dilution
Source
Prediluted Dako 1:200
R&D Systems, Inc.
DAK-Calret Prediluted Dako 1 1A4 Prediluted Dako
Pretreatment
Detection method
TRS high pH
Dako EnVision Flex+ Mouse TRS low pH in PC Dako EnVision+ Mouse Dako PT module high Dako EnVision+ pH Mouse TRS high pH Dako Flex
Not 1:200 Cell Signaling Technology, EDTA high pH applicable Inc. MIB-1 Prediluted Dako TRS high pH
Dako EnVision+ Rabbit Dako Envision Flex
124
Dako Envision Flex
Prediluted Dako
TRS high pH
BCL-2 indicates B-cell lymphoma 2 oncoprotein; PC, pressure cooker; PT, pretreatment; SF-1, steroidogenic factor-1; TRS, target retrieval solution (Dako, Carpinteria, CA).
Int J Gynecol Pathol Vol. 33, No. 1, January 2014
INVOLUTING LUTEINIZED THECOMA OF THE OVARY
25
FIG. 1. Involuting luteinized thecoma. (A) A paucicellular, hyalinized area of the neoplasm resembles a corpus albicans because of its convoluted surface that separates it from cellular ovarian cortical stroma noted in the left upper corner of the field (H&E: 200 ). (B) The neoplasm shows anastomosing hyaline bands separating clusters of plump theca cells. Note the adipose tissue in the left upper portion of the field (H&E: 100 ). (C) Hyaline bands surround theca cell elements (H&E: 400 ). (D) Lutein cell nodule containing large mature adipocytes is surrounded by hyaline connective tissue (H&E: 400 ). (E) Adipose tissue contains residual theca cell elements (H&E: 100 ). (F) A thin band of hyalinized collagen surrounds a paucicellular myxohyaline nodule (H&E: 40 ). H&E indicates hematoxylin and eosin.
inhibin (Fig. 2E) and nuclear staining for SF-1 (Fig. 2F); however, the adipocytes were negative. SMA and calretinin also were positive in the theca
and lutein cells, but the staining was less specific. Cleaved caspase-3, BCL-2, and Ki-67 were negative within tumor cells. Controls stained appropriately. Int J Gynecol Pathol Vol. 33, No. 1, January 2014
26
L.M. ROTH ET AL.
FIG. 2. Involuting luteinized thecoma. (A) The neoplasm contains foci of dystrophic calcification and adipose tissue with compressed ovarian stroma in the lower part of the field (H&E: 40 ). (B) Area of residual theca cells in the tumor shows a focus of calcification occurring in a hyaline band. Residual ovarian stroma is noted at the lower left (H&E: 200 ). (C) Residual foci of theca cells are surrounded by hyalinized connective tissue and large adipocytes (H&E: 100 ). (D) Within residual theca cells, a small cell cluster has pyknotic nuclei and abundant vacuolated cytoplasm (H&E: 400 ). (E) The cytoplasm of the theca cells is positive for inhibin (inhibin: 100 ). (F) A cluster of neoplastic lutein cells shows positive nuclear staining for steroidogenic factor 1 (SF-1: 600 ). H&E indicates hematoxylin and eosin.
Inhibin and SF-1 were also strongly positive in the granulosa-lutein and theca-lutein cell layers of a nearby active corpus luteum; however, the Int J Gynecol Pathol Vol. 33, No. 1, January 2014
staining was more intense in the granulosa-lutein cells. Calretinin stained the theca-lutein cells but was negative in the granulosa-lutein cells. SMA
INVOLUTING LUTEINIZED THECOMA OF THE OVARY was negative in both layers except for vascular staining. No other type of ovarian stromal tumor was identified in the right ovary. The left ovary and both fallopian tubes were unremarkable. The endometrium showed a late secretory phase. No leiomyomas were identified in the myometrium. DISCUSSION Clinically, the patient had dysfunctional bleeding characterized by long-standing painful heavy periods and requested to undergo a hysterectomy. No anatomic cause for the bleeding was found. Given the small size and histologic appearance of the right ovarian neoplasm with only a limited amount of viable theca and lutein cells, we consider it highly unlikely that the tumor was producing clinically significant quantities of steroid hormones. The luteinized thecoma described in our case showed marked histologic evidence of involution. The convoluted structure noted histologically is because of the loss of volume of the original tumor with resultant enfolding of the hyalinized collagen near the border and is similar to that seen in a corpus albicans. Foci of residual theca and lutein cells enabled us to identify the nature of the original neoplasm. Scattered cells with pyknotic nuclei and vacuolated cytoplasm were identified in these foci, suggesting that the process of involution was ongoing but occurring at a slow pace. Fibrosis and hyalinization were prominent features, and dystrophic calcification also occurred. We observed a single nodular deposit of myxohyaline material within the tumor. Mature adipocytes were identified in a nodule of remaining lutein cells, and adipose metaplasia was prominent within the neoplasm. The adipose metaplasia may be related to involution of lutein and theca cells in the tumor. Extensive adipose metaplasia was noted previously in an extensively calcified thecoma that occurred in a young woman (5). Adipose tissue has rarely been observed in the normal ovary and has been referred to as adipose prosoplasia or metaplasia (6). Positive immunohistochemical staining patterns for inhibin and SF-1 within residual theca cells and a lutein cell nodule confirmed the ovarian stromal derivation of the neoplasm. The markers are complementary as inhibin is a cytoplasmic stain and SF-1 stains the nucleus. Calretinin and SMA were also positive. The tumor was negative for the apoptotic marker, cleaved (activated) caspase-3, the
27
anti-apotoptic protein BCL-2, and the proliferation marker Ki-67. Spontaneous tumor involution in ovarian stromal and sex cord-stromal neoplasms is a little known phenomenon that is rarely discussed in the literature. This term has been applied previously to describe both the histologic conversion of the corpus luteum of menstruation to the corpus albicans and the spontaneous decrease in size of some untreated acoustic schwannomas observed over prolonged periods of time by modern imaging techniques (3,4). We are not aware of a histologic definition of spontaneous tumor involution in ovarian stromal and sex cord-stromal tumors in the literature. Therefore, we preliminarily define involution histologically as those changes that are similar to those observed in the physiological conversion of a corpus luteum of menstruation to a corpus albicans. Gilks and Clement (3) described involutional changes in granulosa-lutein cells of the corpus luteum of menstruation that begin on the eighth or ninth day after ovulation consisting of decrease in cell size, nuclear pyknosis, accumulation of abundant cytoplasmic lipid, and eventual cellular dissolution. There is progressive fibrosis and shrinkage over a period of several months and eventual conversion to a corpus albicans. Occasionally, focal calcification or ossification is encountered in corpora albicantia. We use the term involution in our case for 3 reasons. It is a broader term compared with regression, it does not require the identification of apoptosis, and it incorporates a number of changes observed that are not included in the term regression, particularly ischemic and metaplastic changes. The question arose in our study as to whether the process occurring in the neoplasm we are reporting was regression. In her Atlas of Ovarian Tumors published in 1943, Gemma Barzilai (7) illustrated at high-magnification ‘‘regressive’’ features within a thecoma that she compared with the transition of the theca interna of a corpus luteum to the corpus albicans. Nuclei were irregular and pyknotic, and the cytoplasm was vacuolated. Large hyaline bands and aggregates were also described in the tumor but were not illustrated. Since that time, we are aware of only 2 references in the literature to her observation, and both articles doubted that the changes Barzilai observed were truly regressive in nature, partly because of the young age of most of the patients in the cases they were describing (8,9). More recent molecular studies have come to equate the term regression in corpora lutea and in malignant Int J Gynecol Pathol Vol. 33, No. 1, January 2014
28
L.M. ROTH ET AL.
tumors with the identification of apoptosis. Although no information is available in the more recent literature on regression of ovarian stromal tumors, regression of the corpus luteum with conversion to a corpus albicans has been studied extensively in animals, and Stocco et al. (10) have reviewed the topic. There are also a few studies on regression of the human corpus luteum of menstruation (11,12). Other terms that have been applied to the histologic changes in the corpus luteum during its conversion to a corpus albicans include luteolysis or involution. Apoptosis is an important component of the regressive process and is produced by a cascade of molecular steps that are initiated in distinct pathways that culminate in the activation of caspases (13). The histologic counterpart of apoptosis is the identification of apoptotic bodies; however, their nature must be confirmed by immunohistochemical staining patterns using caspase-3 or the TUNEL method. Cleaved (activated) caspase-3 is arguably the most direct and robust way to demonstrate apoptosis immunocytochemically and has been reported to be superior to the TUNEL method in regard to ease of use, sensitivity, and reliability (14). Although we identified cells with pyknotic nuclei and abundant cytoplasmic lipid among lutein and theca cells within the neoplasm we are reporting, we failed to demonstrate apoptotic bodies using cleaved caspase-3 immunostains. Thus, we were unable to confirm the possibility that apoptosis played a significant role in our neoplasm. Because of the absence of apoptotic bodies, we consider the less specific designation, involution to be appropriate for our case. As ovarian stromal and sex cord-stromal tumors differ histologically from each other and from the corpus luteum of menstruation, the process of involution in neoplasms would be expected to differ significantly in details from the physiological process. Although, there is a histologic resemblance of involution in thecomas to the physiological involution of a corpus luteum, there are also significant differences between the 2 processes. Involution of a corpus luteum after menstruation occurs rapidly because of the decrease in the level of luteinizing hormone secreted by the pituitary gland, whereas the factors involved in the involution of thecoma are largely unknown. In benign ovarian stromal tumors, involution likely occurs over a time period of years; thus, the levels of apoptotic markers would be expected to be low, if present at all. At the time we examined the tumor histologically, there was evidence of nuclear pyknosis and vacuolated cytoplasm within Int J Gynecol Pathol Vol. 33, No. 1, January 2014
the residual theca and lutein cells. The nuclei of the great majority of residual theca and lutein cells, however, were vesicular with bland chromatin. Localized ischemic changes likely account for the progressive fibrosis, hyalinization, and dystrophic calcification. Proliferation of the hyaline bands described in thecomas and fibromas may cause localized ischemia, and the latter may lead to further hyalinization and dystrophic calcification. Patchy dystrophic calcification was a minor component of the neoplasm we describe. The calcification appeared to arise within hyalinized collagen bands. The adipose tissue may be derived partly from the theca and lutein cells and is considered to be a type of metaplasia. Extensive adipose metaplasia was noted previously in an extensively calcified thecoma that occurred in a young woman (5). Adipose tissue has rarely been observed in the normal ovary and has been referred to as adipose prosoplasia or metaplasia (6). There are a few reports in the literature that may represent examples of involution in ovarian stromal tumors. Young et al. (5) described 4 cases of extensively calcified thecoma that occurred in young women. In another study, Young et al. (15) found that 11% of juvenile granulosa cell tumors showed focal sclerosis that imparted a superficial resemblance to sclerosing stromal tumor. The differential diagnosis in this case is limited and includes the possibility of a corpus luteum undergoing conversion to a corpus albicans or an involuting aberrant graafian follicle. Regarding the possibility of an involuting corpus luteum of menstruation that was being converted into a corpus albicans, a layering of granulosa- and theca-lutein cells likely would be observed in the earlier stages. The granulosa-lutein cells would have pyknotic nuclei and abundant cytoplasmic lipid. Altered blood, hemosiderin, or lipofuscin pigment might also be seen in the early stages. As a corpus luteum of 1-week age by history and histologic appearance occurred in the same ovary, an involuting corpus luteum under ordinary circumstances would have to be at least 4 to 5 weeks old. The neoplasm we describe was slightly larger on macroscopic examination compared with the accompanying clinically 1-week-old corpus luteum; thus, we consider the possibility of involution of a corpus luteum highly unlikely. Adipose tissue or persistence of functioning theca or lutein cells has not been described in involuting corpora lutea, and hyalinization and calcification would be highly unlikely until a corpus albicans is formed.
INVOLUTING LUTEINIZED THECOMA OF THE OVARY A remote aberrant, but otherwise physiological, follicle could undergo a series of partial involutional changes but follicular granulosa cells would be expected, in addition to or rather than luteinized theca cells. In fact, proliferations of persistent granulosa cells within the center of atretic follicles of pregnant and less commonly nonpregnant women may mimic small granulosa cell tumors, or rarely, Sertoli cell tumors (16). These proliferations have a sex cord component, often are surrounded by luteinized theca interna cells and are microscopic findings measuring up to 5 mm in diameter, much smaller than the neoplasm described in our reported case. In summary, we report a case of luteinized thecoma that showed extensive and variegated evidence of involution. The significance of our case is 2-fold. First, it provides further evidence that spontaneous involution occurs in ovarian stromal tumors. Second, it suggests that the physiological involution of a corpus luteum to a corpus albicans has a neoplastic counterpart in unusual cases of thecoma. Spontaneous tumor involution in ovarian stromal tumors is an unusual and poorly understood phenomenon. REFERENCES 1. Zhang J, Young RH, Arseneau J, et al. Ovarian stromal tumors containing lutein or Leydig cells (luteinized thecomas and stromal Leydig cell tumors)—a clinicopathological analysis of fifty cases. Int J Gynecol Pathol 1982;1:270–85. 2. Roth LM, Sternberg WH. Partly luteinized theca cell tumor of the ovary. Cancer 1983;51:1697–704.
29
3. Gilks CB, Clement PB. Ovary. In: Mills SE, ed. Histopathology for Pathologists 4th ed. Philadelphia: Lippincottt Williams & Wilkins; 2012:1119–48. 4. Luetje CM. Spontaneous involution of acoustic tumors. Am J Otol 2000;21:393–8. 5. Young RH, Clement PB, Scully RE. Calcified thecomas in young women. A report of four cases. Int J Gynecol Pathol 1988;7:343–50. 6. Hart WR, Abell MR. Adipose prosoplasia of ovary. Am J Obstet Gynecol 1970;106:929–31. 7. Barzilai G. Atlas of Ovarian Tumors. New York: Grune & Stratton; 1943. 8. Chalvardjian A, Scully RE. Sclerosing stromal tumors of the ovary. Cancer 1973;31:664–70. 9. Ramzy I. Signet-ring stromal tumor of ovary. Histochemical, light, and electron microscopic study. Cancer 1976;38:166–72. 10. Stocco C, Telleria C, Gibori G. The molecular control of corpus luteum formation, function, and regression. Endocr Rev 2007;28:117–49. 11. Vaskivuo TE, Ottander U, Oduwole O, et al. Role of apoptosis, apoptosis-related factors and 17beta-hydroxysteroid dehydrogenases in human corpus luteum regression. Mol Cell Endocrinol 2002;194:191–200. 12. Vega M, Urrutia L, Iniguez G, et al. Nitric oxide induces apoptosis in the human corpus luteum in vitro. Mol Hum Reprod 2000;6:681–7. 13. Bishop EF, Badve S, Morimiya A, et al. Apoptosis in spermatocytic and usual seminomas: a light microscopic and immunohistochemical study. Mod Pathol 2007;20:1036–44. 14. Duan WR, Garner DS, Williams SD, et al. Comparison of immunohistochemistry for activated caspase-3 and cleaved cytokeratin 18 with the TUNEL method for quantification of apoptosis in histological sections of PC-3 subcutaneous xenografts. J Pathol 2003;199:221–8. 15. Young RH, Dickersin GR, Scully RE. Juvenile granulosa cell tumor of the ovary. A clinicopathological analysis of 125 cases. Am J Surg Pathol 1984;8:575–96. 16. Clement PB, Young RH, Scully RE. Ovarian granulosa cell proliferations of pregnancy: a report of nine cases. Hum Pathol 1988;19:657–62.
Int J Gynecol Pathol Vol. 33, No. 1, January 2014
Case Report
Involuting Luteinized Thecoma of the Ovary Lawrence M. Roth, M.D., Arthur R. Gaba, M.D., Michael D. Linden, and George E. Sandusky, D.V.M., Ph.D.
M.D.,
Summary: Spontaneous tumor involution in ovarian stromal tumors is a poorly understood phenomenon. In this report, we describe a rare case of luteinized thecoma that showed extensive involutional changes, such that cellular elements diagnostic of luteinized thecoma were sparse. The convoluted contour of the tumor resembled that observed in a corpus albicans; however, the neoplasm was considerably larger, and the contents of the nodule differed from that of a corpus albicans. The diagnosis of luteinized thecoma was established by the identification of residual aggregates of neoplastic theca cells and a nodule of lutein cells that were positive for inhibin and steroidogenic factor-1. Features of involution within the tumor included a few theca and lutein cells with pyknotic nuclei and abundant cytoplasmic lipid, occasional large adipocytes among the lutein cells, extensive hyalinization, dystrophic calcification, a myxohyaline nodule, and adipose metaplasia. It is likely that some of the aforementioned changes are the result of accompanying ischemia. Cleaved caspase-3 staining patterns were negative within residual lutein and theca cells; thus, we were unable to establish the occurrence of apoptotic bodies. Key Words: Ovary—Luteinized thecoma—Involution.
histologic changes that occur during the conversion of the corpus luteum to a corpus albicans and to those untreated unilateral acoustic schwannomas that decreased in size when followed for extended time periods by imaging techniques (3,4).
Luteinized thecoma is an uncommon, usually benign ovarian stromal tumor. In 1982, Zhang et al. (1) reported the largest series of 46 cases, 4 of which were malignant. The following year, 4 additional benign cases were described (2). Luteinized thecomas occur in younger women more often compared with typical thecomas, and 30% arise in women below 30 years of age (1). In this article, we use the term involution to describe secondary changes that occurred in the neoplasm we are reporting. This term has been applied previously to describe the
MATERIALS AND METHODS The tissue blocks were fixed in formalin, routinely processed, and embedded in paraffin. The slides were deparaffinized twice in xylene for 5 minutes, rehydrated using graded ethanol solutions to distilled water, and stained with hematoxylin and eosin. All slides were retrospectively reviewed and diagnosed according to the established criteria. Four-mm-thick sections were cut for immunohistochemical staining that was performed on a Dako Autostainer Plus (Dako, Carpinteria, CA). In brief, endogenous peroxidase activity was blocked using 3% hydrogen peroxide in methanol, and indigenous biotin activity was blocked using a protein blocker.
From the Departments of Pathology, Indiana University School of Medicine (L.M.R., G.E.S.), Indianapolis, Indiana; Henry Ford Health System (A.R.G.), Detroit, Michigan; and University of California San Diego School of Medicine and Veterans Administration San Diego Healthcare System (M.D.L.), San Diego, California. The authors declare no conflict of interest. Address correspondence and reprint requests to Lawrence M. Roth, MD, Department of Pathology, Indiana University School of Medicine, Van Nuys Medical Science Building 128, 635 Barnhill Drive, Indianapolis, IN 46202-5120. E-mail: [email protected].
23
DOI: 10.1097/PGP.0b013e31827db06f
24
L.M. ROTH ET AL.
After antigen retrieval, the slides were incubated with primary antibody. The detection systems used were Dako EnVision+, EnVision FLEX, and EnVision FLEX+. The peroxidase activity was developed with 3,3-diaminobenzidine and counterstained with hematoxylin. Appropriate positive controls for each antibody were run concurrently and showed adequate immunostaining. Slides of the neoplasm and an adjacent corpus luteum were analyzed for immunoreactivity using the following antibodies: inhibin a (inhibin), steroidogenic factor-1 (SF-1), calretinin, a-smooth muscle actin (SMA), cleaved caspase-3, Ki-67, and B-cell lymphoma 2 oncoprotein (BCL-2). Table 1 indicates the types of antibody used, the clone used for monoclonal antibodies, dilution, source, pretreatment, and the detection method. CASE REPORT A 49-year-old woman presented with a history of heavy, painful periods. She was G2 P1001. Her last menstrual period occurred 1 week before presentation. Transabdominal and endovaginal imaging of the pelvis showed a retroflexed uterus that measured 9 3.4 4.7 cm. The endometrial stripe measured 6.8 mm. The uterus and ovaries were unremarkable. The impression was an unremarkable pelvic ultrasound. At the patient’s request, a robotic-assisted hysterectomy and a bilateral salpingo-oophorectomy were performed for the relief of symptoms. There was no evidence of disease 1 month after the surgery. The specimen obtained from the surgery included the right ovary that measured 3.5 2.5 2.0 cm and an attached fallopian tube that measured 6 cm in
length. The external surface of the ovary was yellowtan in color and cerebriform in appearance. The sectioned surface showed a well-circumscribed tanwhite and yellow nodule measuring 1.7 1.4 1.3 cm. In addition, a hemorrhagic, cystic corpus luteum was identified that measured 1.5 1.5 1.1 cm. The uterus weighed 115.5 g, measured 8.0 6.5 4.2 cm, and showed no abnormalities. The left ovary measured 2.5 2.0 1.0 cm, and the fallopian tube measured 6.5 cm in length. None of these showed any abnormalities. Histologically, the right ovarian neoplasm was sharply demarcated from the surrounding ovarian stroma, and the surface of the neoplasm had a convoluted appearance that mimicked a corpus albicans (Fig. 1A). In some areas, aggregates of plump, spindle-shaped cells with elongated nuclei resembling thecoma were surrounded by hyalinized connective tissue (Figs. 1B, C). A nodule of lutein cells containing several large adipocytes was enveloped by hyalinized connective tissue (Fig. 1D). Adipose tissue was prominent and contained occasional clusters of residual theca cells (Fig. 1E). A thin band of hyaline connective tissue surrounded a single myxohyaline nodule (Fig. 1F). The tumor was sharply circumscribed and compressed the surrounding ovarian stroma (Fig. 2A). Patchy dystrophic calcification sometimes occurred within hyaline bands in association with involuting theca cells (Fig. 2B). Residual clusters of theca cells were surrounded by hyalinized connective tissue and adipocytes (Fig. 2C). Most theca cells had bland vesicular nuclei, but some had pyknotic nuclei and abundant vacuolated cytoplasm (Fig. 2D). The theca and lutein cells showed cytoplasmic staining for
TABLE 1. Details for antibodies and conditions used in this study Antibody Inhibin a SF-1 Calretinin a-Smooth muscle actin Cleaved caspase-3 Ki-67 BCL-2
Type Mouse monoclonal Mouse monoclonal Mouse monoclonal Mouse monoclonal Rabbit polyclonal Mouse monoclonal Mouse monoclonal
Clone R1 N1665
Dilution
Source
Prediluted Dako 1:200
R&D Systems, Inc.
DAK-Calret Prediluted Dako 1 1A4 Prediluted Dako
Pretreatment
Detection method
TRS high pH
Dako EnVision Flex+ Mouse TRS low pH in PC Dako EnVision+ Mouse Dako PT module high Dako EnVision+ pH Mouse TRS high pH Dako Flex
Not 1:200 Cell Signaling Technology, EDTA high pH applicable Inc. MIB-1 Prediluted Dako TRS high pH
Dako EnVision+ Rabbit Dako Envision Flex
124
Dako Envision Flex
Prediluted Dako
TRS high pH
BCL-2 indicates B-cell lymphoma 2 oncoprotein; PC, pressure cooker; PT, pretreatment; SF-1, steroidogenic factor-1; TRS, target retrieval solution (Dako, Carpinteria, CA).
Int J Gynecol Pathol Vol. 33, No. 1, January 2014
INVOLUTING LUTEINIZED THECOMA OF THE OVARY
25
FIG. 1. Involuting luteinized thecoma. (A) A paucicellular, hyalinized area of the neoplasm resembles a corpus albicans because of its convoluted surface that separates it from cellular ovarian cortical stroma noted in the left upper corner of the field (H&E: 200 ). (B) The neoplasm shows anastomosing hyaline bands separating clusters of plump theca cells. Note the adipose tissue in the left upper portion of the field (H&E: 100 ). (C) Hyaline bands surround theca cell elements (H&E: 400 ). (D) Lutein cell nodule containing large mature adipocytes is surrounded by hyaline connective tissue (H&E: 400 ). (E) Adipose tissue contains residual theca cell elements (H&E: 100 ). (F) A thin band of hyalinized collagen surrounds a paucicellular myxohyaline nodule (H&E: 40 ). H&E indicates hematoxylin and eosin.
inhibin (Fig. 2E) and nuclear staining for SF-1 (Fig. 2F); however, the adipocytes were negative. SMA and calretinin also were positive in the theca
and lutein cells, but the staining was less specific. Cleaved caspase-3, BCL-2, and Ki-67 were negative within tumor cells. Controls stained appropriately. Int J Gynecol Pathol Vol. 33, No. 1, January 2014
26
L.M. ROTH ET AL.
FIG. 2. Involuting luteinized thecoma. (A) The neoplasm contains foci of dystrophic calcification and adipose tissue with compressed ovarian stroma in the lower part of the field (H&E: 40 ). (B) Area of residual theca cells in the tumor shows a focus of calcification occurring in a hyaline band. Residual ovarian stroma is noted at the lower left (H&E: 200 ). (C) Residual foci of theca cells are surrounded by hyalinized connective tissue and large adipocytes (H&E: 100 ). (D) Within residual theca cells, a small cell cluster has pyknotic nuclei and abundant vacuolated cytoplasm (H&E: 400 ). (E) The cytoplasm of the theca cells is positive for inhibin (inhibin: 100 ). (F) A cluster of neoplastic lutein cells shows positive nuclear staining for steroidogenic factor 1 (SF-1: 600 ). H&E indicates hematoxylin and eosin.
Inhibin and SF-1 were also strongly positive in the granulosa-lutein and theca-lutein cell layers of a nearby active corpus luteum; however, the Int J Gynecol Pathol Vol. 33, No. 1, January 2014
staining was more intense in the granulosa-lutein cells. Calretinin stained the theca-lutein cells but was negative in the granulosa-lutein cells. SMA
INVOLUTING LUTEINIZED THECOMA OF THE OVARY was negative in both layers except for vascular staining. No other type of ovarian stromal tumor was identified in the right ovary. The left ovary and both fallopian tubes were unremarkable. The endometrium showed a late secretory phase. No leiomyomas were identified in the myometrium. DISCUSSION Clinically, the patient had dysfunctional bleeding characterized by long-standing painful heavy periods and requested to undergo a hysterectomy. No anatomic cause for the bleeding was found. Given the small size and histologic appearance of the right ovarian neoplasm with only a limited amount of viable theca and lutein cells, we consider it highly unlikely that the tumor was producing clinically significant quantities of steroid hormones. The luteinized thecoma described in our case showed marked histologic evidence of involution. The convoluted structure noted histologically is because of the loss of volume of the original tumor with resultant enfolding of the hyalinized collagen near the border and is similar to that seen in a corpus albicans. Foci of residual theca and lutein cells enabled us to identify the nature of the original neoplasm. Scattered cells with pyknotic nuclei and vacuolated cytoplasm were identified in these foci, suggesting that the process of involution was ongoing but occurring at a slow pace. Fibrosis and hyalinization were prominent features, and dystrophic calcification also occurred. We observed a single nodular deposit of myxohyaline material within the tumor. Mature adipocytes were identified in a nodule of remaining lutein cells, and adipose metaplasia was prominent within the neoplasm. The adipose metaplasia may be related to involution of lutein and theca cells in the tumor. Extensive adipose metaplasia was noted previously in an extensively calcified thecoma that occurred in a young woman (5). Adipose tissue has rarely been observed in the normal ovary and has been referred to as adipose prosoplasia or metaplasia (6). Positive immunohistochemical staining patterns for inhibin and SF-1 within residual theca cells and a lutein cell nodule confirmed the ovarian stromal derivation of the neoplasm. The markers are complementary as inhibin is a cytoplasmic stain and SF-1 stains the nucleus. Calretinin and SMA were also positive. The tumor was negative for the apoptotic marker, cleaved (activated) caspase-3, the
27
anti-apotoptic protein BCL-2, and the proliferation marker Ki-67. Spontaneous tumor involution in ovarian stromal and sex cord-stromal neoplasms is a little known phenomenon that is rarely discussed in the literature. This term has been applied previously to describe both the histologic conversion of the corpus luteum of menstruation to the corpus albicans and the spontaneous decrease in size of some untreated acoustic schwannomas observed over prolonged periods of time by modern imaging techniques (3,4). We are not aware of a histologic definition of spontaneous tumor involution in ovarian stromal and sex cord-stromal tumors in the literature. Therefore, we preliminarily define involution histologically as those changes that are similar to those observed in the physiological conversion of a corpus luteum of menstruation to a corpus albicans. Gilks and Clement (3) described involutional changes in granulosa-lutein cells of the corpus luteum of menstruation that begin on the eighth or ninth day after ovulation consisting of decrease in cell size, nuclear pyknosis, accumulation of abundant cytoplasmic lipid, and eventual cellular dissolution. There is progressive fibrosis and shrinkage over a period of several months and eventual conversion to a corpus albicans. Occasionally, focal calcification or ossification is encountered in corpora albicantia. We use the term involution in our case for 3 reasons. It is a broader term compared with regression, it does not require the identification of apoptosis, and it incorporates a number of changes observed that are not included in the term regression, particularly ischemic and metaplastic changes. The question arose in our study as to whether the process occurring in the neoplasm we are reporting was regression. In her Atlas of Ovarian Tumors published in 1943, Gemma Barzilai (7) illustrated at high-magnification ‘‘regressive’’ features within a thecoma that she compared with the transition of the theca interna of a corpus luteum to the corpus albicans. Nuclei were irregular and pyknotic, and the cytoplasm was vacuolated. Large hyaline bands and aggregates were also described in the tumor but were not illustrated. Since that time, we are aware of only 2 references in the literature to her observation, and both articles doubted that the changes Barzilai observed were truly regressive in nature, partly because of the young age of most of the patients in the cases they were describing (8,9). More recent molecular studies have come to equate the term regression in corpora lutea and in malignant Int J Gynecol Pathol Vol. 33, No. 1, January 2014
28
L.M. ROTH ET AL.
tumors with the identification of apoptosis. Although no information is available in the more recent literature on regression of ovarian stromal tumors, regression of the corpus luteum with conversion to a corpus albicans has been studied extensively in animals, and Stocco et al. (10) have reviewed the topic. There are also a few studies on regression of the human corpus luteum of menstruation (11,12). Other terms that have been applied to the histologic changes in the corpus luteum during its conversion to a corpus albicans include luteolysis or involution. Apoptosis is an important component of the regressive process and is produced by a cascade of molecular steps that are initiated in distinct pathways that culminate in the activation of caspases (13). The histologic counterpart of apoptosis is the identification of apoptotic bodies; however, their nature must be confirmed by immunohistochemical staining patterns using caspase-3 or the TUNEL method. Cleaved (activated) caspase-3 is arguably the most direct and robust way to demonstrate apoptosis immunocytochemically and has been reported to be superior to the TUNEL method in regard to ease of use, sensitivity, and reliability (14). Although we identified cells with pyknotic nuclei and abundant cytoplasmic lipid among lutein and theca cells within the neoplasm we are reporting, we failed to demonstrate apoptotic bodies using cleaved caspase-3 immunostains. Thus, we were unable to confirm the possibility that apoptosis played a significant role in our neoplasm. Because of the absence of apoptotic bodies, we consider the less specific designation, involution to be appropriate for our case. As ovarian stromal and sex cord-stromal tumors differ histologically from each other and from the corpus luteum of menstruation, the process of involution in neoplasms would be expected to differ significantly in details from the physiological process. Although, there is a histologic resemblance of involution in thecomas to the physiological involution of a corpus luteum, there are also significant differences between the 2 processes. Involution of a corpus luteum after menstruation occurs rapidly because of the decrease in the level of luteinizing hormone secreted by the pituitary gland, whereas the factors involved in the involution of thecoma are largely unknown. In benign ovarian stromal tumors, involution likely occurs over a time period of years; thus, the levels of apoptotic markers would be expected to be low, if present at all. At the time we examined the tumor histologically, there was evidence of nuclear pyknosis and vacuolated cytoplasm within Int J Gynecol Pathol Vol. 33, No. 1, January 2014
the residual theca and lutein cells. The nuclei of the great majority of residual theca and lutein cells, however, were vesicular with bland chromatin. Localized ischemic changes likely account for the progressive fibrosis, hyalinization, and dystrophic calcification. Proliferation of the hyaline bands described in thecomas and fibromas may cause localized ischemia, and the latter may lead to further hyalinization and dystrophic calcification. Patchy dystrophic calcification was a minor component of the neoplasm we describe. The calcification appeared to arise within hyalinized collagen bands. The adipose tissue may be derived partly from the theca and lutein cells and is considered to be a type of metaplasia. Extensive adipose metaplasia was noted previously in an extensively calcified thecoma that occurred in a young woman (5). Adipose tissue has rarely been observed in the normal ovary and has been referred to as adipose prosoplasia or metaplasia (6). There are a few reports in the literature that may represent examples of involution in ovarian stromal tumors. Young et al. (5) described 4 cases of extensively calcified thecoma that occurred in young women. In another study, Young et al. (15) found that 11% of juvenile granulosa cell tumors showed focal sclerosis that imparted a superficial resemblance to sclerosing stromal tumor. The differential diagnosis in this case is limited and includes the possibility of a corpus luteum undergoing conversion to a corpus albicans or an involuting aberrant graafian follicle. Regarding the possibility of an involuting corpus luteum of menstruation that was being converted into a corpus albicans, a layering of granulosa- and theca-lutein cells likely would be observed in the earlier stages. The granulosa-lutein cells would have pyknotic nuclei and abundant cytoplasmic lipid. Altered blood, hemosiderin, or lipofuscin pigment might also be seen in the early stages. As a corpus luteum of 1-week age by history and histologic appearance occurred in the same ovary, an involuting corpus luteum under ordinary circumstances would have to be at least 4 to 5 weeks old. The neoplasm we describe was slightly larger on macroscopic examination compared with the accompanying clinically 1-week-old corpus luteum; thus, we consider the possibility of involution of a corpus luteum highly unlikely. Adipose tissue or persistence of functioning theca or lutein cells has not been described in involuting corpora lutea, and hyalinization and calcification would be highly unlikely until a corpus albicans is formed.
INVOLUTING LUTEINIZED THECOMA OF THE OVARY A remote aberrant, but otherwise physiological, follicle could undergo a series of partial involutional changes but follicular granulosa cells would be expected, in addition to or rather than luteinized theca cells. In fact, proliferations of persistent granulosa cells within the center of atretic follicles of pregnant and less commonly nonpregnant women may mimic small granulosa cell tumors, or rarely, Sertoli cell tumors (16). These proliferations have a sex cord component, often are surrounded by luteinized theca interna cells and are microscopic findings measuring up to 5 mm in diameter, much smaller than the neoplasm described in our reported case. In summary, we report a case of luteinized thecoma that showed extensive and variegated evidence of involution. The significance of our case is 2-fold. First, it provides further evidence that spontaneous involution occurs in ovarian stromal tumors. Second, it suggests that the physiological involution of a corpus luteum to a corpus albicans has a neoplastic counterpart in unusual cases of thecoma. Spontaneous tumor involution in ovarian stromal tumors is an unusual and poorly understood phenomenon. REFERENCES 1. Zhang J, Young RH, Arseneau J, et al. Ovarian stromal tumors containing lutein or Leydig cells (luteinized thecomas and stromal Leydig cell tumors)—a clinicopathological analysis of fifty cases. Int J Gynecol Pathol 1982;1:270–85. 2. Roth LM, Sternberg WH. Partly luteinized theca cell tumor of the ovary. Cancer 1983;51:1697–704.
29
3. Gilks CB, Clement PB. Ovary. In: Mills SE, ed. Histopathology for Pathologists 4th ed. Philadelphia: Lippincottt Williams & Wilkins; 2012:1119–48. 4. Luetje CM. Spontaneous involution of acoustic tumors. Am J Otol 2000;21:393–8. 5. Young RH, Clement PB, Scully RE. Calcified thecomas in young women. A report of four cases. Int J Gynecol Pathol 1988;7:343–50. 6. Hart WR, Abell MR. Adipose prosoplasia of ovary. Am J Obstet Gynecol 1970;106:929–31. 7. Barzilai G. Atlas of Ovarian Tumors. New York: Grune & Stratton; 1943. 8. Chalvardjian A, Scully RE. Sclerosing stromal tumors of the ovary. Cancer 1973;31:664–70. 9. Ramzy I. Signet-ring stromal tumor of ovary. Histochemical, light, and electron microscopic study. Cancer 1976;38:166–72. 10. Stocco C, Telleria C, Gibori G. The molecular control of corpus luteum formation, function, and regression. Endocr Rev 2007;28:117–49. 11. Vaskivuo TE, Ottander U, Oduwole O, et al. Role of apoptosis, apoptosis-related factors and 17beta-hydroxysteroid dehydrogenases in human corpus luteum regression. Mol Cell Endocrinol 2002;194:191–200. 12. Vega M, Urrutia L, Iniguez G, et al. Nitric oxide induces apoptosis in the human corpus luteum in vitro. Mol Hum Reprod 2000;6:681–7. 13. Bishop EF, Badve S, Morimiya A, et al. Apoptosis in spermatocytic and usual seminomas: a light microscopic and immunohistochemical study. Mod Pathol 2007;20:1036–44. 14. Duan WR, Garner DS, Williams SD, et al. Comparison of immunohistochemistry for activated caspase-3 and cleaved cytokeratin 18 with the TUNEL method for quantification of apoptosis in histological sections of PC-3 subcutaneous xenografts. J Pathol 2003;199:221–8. 15. Young RH, Dickersin GR, Scully RE. Juvenile granulosa cell tumor of the ovary. A clinicopathological analysis of 125 cases. Am J Surg Pathol 1984;8:575–96. 16. Clement PB, Young RH, Scully RE. Ovarian granulosa cell proliferations of pregnancy: a report of nine cases. Hum Pathol 1988;19:657–62.
Int J Gynecol Pathol Vol. 33, No. 1, January 2014
