Search

The aim of this study was to compare survival and recurrence in clinical and surgical stage I-II papillary serous (PS), clear cell (CC), and endometrioid (EM) cancers of the endometrium and examine the prognostic utility of myometrial invasion.Clinical, surgicopathologic, and survival data were retrospectively collected on 574 clinical stage I-II endometrial cancer patients, including 53 PS and 18 CC (based on postoperative histology), undergoing hysterectomy at Duke University Medical Center between 1967 and 1990. All staging material was available and reexamined prior to this analysis, and FIGO surgical staging was retrospectively assigned. Prognostic variables examined included age, stage, grade, myometrial invasion, lymph-vascular space invasion (LVSI), and histology. PS and CC histologic subtypes were compared as both common category and discrete categories versus EM, EM grade 1 (EM1), EM grade 2 (EM2), and EM grade 3 (EM3). Statistical analyses were performed using chi(2), Fisher's exact, and Wilcoxon rank sum tests, Cox regression analysis, and Kaplan-Meier survival analysis.PS tumors accounted for 9%, CC for 3%, and EM for 88% of cases. Recurrences were more frequent among PS (38%) and CC (22%) compared with EM (9%) (P0.001 and 0.08, respectively), and PS recurred more frequently than EM3 alone (20%) (P = 0.06). Among PS, CC, and EM3 patients with recurrences there were no statistical differences in the proportion that received preoperative or postoperative radiotherapy or chemotherapy. Prognostic factors for shorter survival included age=60, surgical stage III+IV, presence of LVSI, histology (PS, CC, or EM3), and=50% myometrial invasion. The estimated 5-year survival of PS+CC patients with2 mm myometrial invasion is 0.56 compared to 0.93 for EM patients (P0. 001). PS + CC tumors confined to the endometrium had a 5-year survival of 0.60 compared to 0.98 and 1.00 for EM and EM3, respectively. The 5-year survival for surgically staged IA patients (0.57) was not different from stages IB and IC combined (0.53) (P = 0.72). The 5-year survival for surgical stage I + II PS + CC patients (0.56) was comparable to that for clinical stage I + II PS + CC patients (0.46) and remained significantly smaller than that for EM patients (0.86) (P0.001).Recurrences are more frequent among PS and CC tumors compared with EM and among PS compared with EM3. When controlled for surgical stage I-II tumors, 5-year survival for PS + CC patients remains comparable to that of clinical stage I-II patients and below that of EM. Prognostic factors for survival in PS and CC patients include age, stage, and LVSI. PS, CC, and EM3 subtypes together are predictors of poor survival. Thorough extended surgical staging is indicated in PS and CC tumors, and prospective trials of aggressive adjuvant therapies for surgical stage I-II tumors are needed to improve outcome in PS and CC patients.

Background . Crohn's disease is a chronic inflammatory disorder characterized by focal, transmural inflammation of the intestine. Gynecologic involvement, including rectovaginal fistula formation, is frequent. Case #1 . A 53-year-old female with a 30-year history of Crohn's disease and numerous perirectal fistulas developed a foul smelling, purulent drainage from her rectum and a mucopurulent, bloody discharge from her vagina. A lower vaginal lesion biopsy demonstrated a low-grade mucinous adenocarcinoma. Case #2 . A 42-year-old female with a 15-year history of Crohn's disease developed drainage from her vagina. Physical examination revealed an enlarging mass involving the posterior wall of the vaginal vault that connected to the anus by a fistula tract. A biopsy revealed mucinous adenocarcinoma. Conclusion . Malignant transformation of persistent rectovaginal fistulas is a potential complication of Crohn's disease.

Since overexpression of mutant p53 protein is a common feature of invasive epithelial ovarian cancers, we investigated whether overexpression of the p53 tumor suppressor gene product occurs in benign and borderline epithelial ovarian tumors. Immunohistochemical staining for p53 was performed in frozen samples of 17 benign tumors and in 49 borderline tumors (4 frozen, 45 paraffin embedded). Overexpression of p53 was observed in 0/17 (0%) benign ovarian tumors and 2/49 (4%) borderline tumors. Overexpression of p53 in borderline tumors was only seen in advanced stage cases; overexpression was seen in 2/8 (25%) stage III cases, but not in any of 41 stage I/II cases. In conclusion, overexpression of p53 is not a feature of benign epithelial ovarian tumors or early-stage borderline ovarian tumors. Similar to invasive epithelial ovarian cancers, however, a fraction of metastatic borderline tumors also overexpress p53.

Objective . Placental site trophoblastic tumor (PSTT) is a rare form of gestational trophoblastic disease. Little is known about its pathogenesis and natural history. Methods . This report describes two cases that arose in patients with documented complete hydatidiform moles and summarizes the antecedent prenatal histories of PSTTs based on a detailed Medline literature analysis. Cases . A 28-year-old, G 2 P 2 female had a live, 12-week gestation fetus and a coexisting molar pregnancy. Her hCG levels dropped promptly from 1.5 million to 23,273 IU/ml after termination, but rose shortly thereafter together with the onset of recurrent vaginal bleeding. Curettage revealed persistent mole. Persistently elevated hCG led to hysterectomy disclosing a fundal PSTT. The second case was that of a 48-year-old, G 2 woman who presented with symptoms of preeclampsia, hyperthyroidism, and elevated hCG. Curettage yielded a complete hydatidiform mole. Although the hCG level decreased for a short period, it soon increased despite treatment with methotrexate. A second curettage revealed a PSTT. Discussion . A Medline literature analysis of PSTT, which consists almost entirely of individual cases and several small series, disclosed that PSTT is preceded in 61% of cases by normal term pregnancy, 12% molar pregnancy, 9% spontaneous abortion, 8% therapeutic abortion, and 3% with ectopic pregnancy, stillbirths or preterm delivery. No information is known in 7%. This report describes two additional cases of PSTT preceded by complete molar pregnancy. Conclusions . PSTT is a well recognized, but uncommon form of gestational trophoblastic disease. Although little is known about its pathogenesis, it is preceded not uncommonly by an abnormal pregnancy, including a molar pregnancy.

Objective. Mullerian inclusion cysts (MIC) are small benign appearing glands that are occasionally noted in lymph nodes and peritoneal biopsies. They occur most frequently in women with serous ovarian tumors, with borderline tumors (SBOT) having a higher incidence than invasive cancers. The aim of this study was to examine whether MIC and SBOT have identical K- ras mutations, which would suggest that they are related. Methods. Six patients in whom adequate tissue was available from SBOT, MIC, and normal tissue were identified from a consecutive series of patients with SBOT who underwent lymph node sampling from 1992 to 1997 at Duke University Medical Center. DNA extraction was performed using laser capture microdissection. Exon 1 of the K- ras gene was amplified using PCR and subjected to single-strand conformation analysis to screen for mutations. Shifted bands were sequenced to confirm the presence of mutations. Results. Mutations in codon 12 of K- ras were found in three of six (50%) SBOT. In two of these three cases, the identical mutation was found in the SBOT and the MIC (gly to val in both cases), but not in the corresponding normal DNA. In one case, a mutation was seen in the ovarian tumor (gly to asp), but not in the corresponding MIC. Conclusions. Mutations in codon 12 of the K- ras gene are a hallmark of serous borderline tumors. The presence of identical K- ras mutations in some SBOT and their associated MIC suggests that they are related processes. Both may arise due to a field effect, or alternatively some MIC may represent metastases from the primary ovarian tumor.

Purpose. The aim of this study was to identify similarities and differences in epidemiologic and surgicopathologic staging results for papillary serous (PS) and clear cell (CC) endometrial cancers compared with endometrioid (EM) carcinoma of the endometrium. Methods. Clinical and surgicopathologic data were retrospectively collected on 574 clinical stage I–II endometrial cancer patients, including 53 PS and 18 CC (based on postoperative histology), undergoing hysterectomy at Duke University Medical Center between 1967 and 1990. All staging material was available and reexamined prior to this analysis, and FIGO surgical staging was retrospectively assigned. PS and CC histologic subtypes were compared both as a common category and as discrete categories versus EM, EM grade 1 (EM1), EM grade 2 (EM2), and EM grade 3 (EM3). Fisher's exact test was used to compare proportions with unordered categories (2 × 2 tables), while the χ 2 test for trend was used to compare proportions in 3 × 2 tables with ordered categories. Differences in medians were compared with the Wilcoxon rank-sum test. Results. PS tumors accounted for 8%, CC for 2%, and EM for 90% of cases. Overall, 14% of tumors were changed to a different postoperative histology including 64% of PS, 50% of CC, and 8% of EM. Postoperative histology changes were 4% for EM1 and 21% for EM3. PS, CC, and EM3 had more surgical sampling performed than for other EM. Rates for lymph node dissections were similar for EM3 (81%), PS (72%), and CC (67%) tumors, although metastases were more frequent for PS and CC compared with EM3. When PS tumors were confined to the endometrium, paraaortic metastases occurred in 13%. LVSI increased with EM grade and was highest for PS and CC. Upstaging to surgical stage III–IV occurred in 47% of PS, 39% of CC, and 12% of EM. The majority of PS and CC tumors were confined to the inner one-third of the myometrium, compared with EM tumors, where grade correlated with depth of myometrial invasion. Extrauterine metastases occurred in 55% of PS and 45% of CC tumors confined to the inner one-half, compared with 17% of EM3. Conclusion. Frequent changes from preoperative to postoperative histology and grade may contribute to misassignment of preoperative and intraoperative risk as determined by depth of myometrial invasion for PS and CC patients. The higher frequency of extrauterine metastases in PS and CC tumors compared with EM3, despite similar surgical sampling rates, supports a more virulent behavior. The poor correlation between depth of myometrial invasion and risk for extrauterine metastases helps to explain poorer survival in PS and CC patients, in addition to more frequent upstaging. These results support routine extended surgical staging for women with preoperative or intraoperative diagnosis of PS and CC tumors. Intraoperative assessment of tumor grade and histology may be indicated and warrants further investigation.

Objective: To examine the effect of postsurgical surveillance on survival of patients with FIGO stage I/II endometrial adenocarcinoma. Methods: We examined the records of 354 patients who underwent primary surgical therapy for FIGO stage I/II endometrial adenocarcinoma. In patients who developed recurrent disease, we determined whether symptoms or signs of disease were present at recurrence and whether there was evidence of disease on Pap smear or chest radiograph. Results: Among the 354 patients in this study, 44 (12%) developed recurrent disease. Sites of recurrence included 12 (27%) isolated vaginal, 12 (27%) pelvis with vagina or abdomen, 4 (10%) isolated lung, 13 (29%) pelvic/abdominal with other distant sites, and 3 (7%) other distant sites. At diagnosis of recurrence 61% of patients had symptoms related to their cancer, 68% had physical exam findings suggestive of recurrence, and 84% had symptoms and/or signs. Findings consistent with recurrent cancer were detected by Pap smear in 25% and on chest radiograph in 20%. Among the 44 patients who developed recurrent disease, 8 (18%) remain alive without evidence of disease, including 6/12 (50%) with isolated vaginal disease and 2/34 (6%) with other patterns of recurrent disease ( P = 0.01). Among the 12 patients with isolated vaginal recurrence, 1/3 (33%) in whom recurrent disease was diagnosed by Pap smear alone was salvaged compared to 5/9 (56%) who had symptoms or signs of vaginal recurrence. None of the three patients in whom an abnormal chest radiograph was the only evidence of recurrence survived. Conclusions: Because of the low recurrence rate of FIGO stage I/II endometrial cancer and the paucity of effective second-line treatment, surveillance Pap smears and chest radiographs appear to have little impact on survival. Although few asymptomatic potentially curable recurrences were detected due to surveillance examinations, the value of psychological reassurance associated with a normal examination is difficult to quantitate.

Black women with endometrial cancer have more advanced disease and less favorable tumor grade than do white women. This study evaluated whether racial differences in tumor grade could be explained by hormone-related factors and other putative determinants of grade. Subjects included 207 white and 81 black postmenopausal women diagnosed with primary cancer of the uterine corpus between 1985 and 1987. Blacks had poorer tumor grade than whites (odds ratio for FIGO grade 2 versus grade 1 is 1.8; odds ratio for grade 3 versus grade 1 is 2.8). Over 75% of the excess of poorly differentiated tumors versus well-differentiated tumors among blacks could be explained by racial differences in use of replacement estrogens, age at first pregnancy, history of oophorectomy, poverty, stage of disease, use of screening, and access to health care. The most prominent factor was estrogen therapy, which was associated with favorable tumor grade and was used much less frequently by blacks. Although not statistically significant, a moderate racial difference in tumor grade remained after control of the potential explanatory explanatory variables. This may reflect true biologic variation between blacks and whites and may explain, in part, the observation that blacks with endometrial cancer have a worse prognosis.

Since 1971 several studies have reported the appearance of clear-cell carcinoma and clear-cell cervical carcinoma in young women exposed prenatally to diethylstilbestrol (DES). Several recent reports have been conflicting on this subject. The reported prevalence rate of dysplasia and carcinoma in situ (CIS) range from 0% to 18%, or from 0% to 6% for severe cases. Factors accounting for such discrepancies can be overdiagnosis or a biased selection of patients; as a study of 1424 patients conducted at the Massachusetts General Hospital has shown, the overall rate of severe dysplasia and CIS appeared relatively high only until patients were divided into 2 groups, those referred because of earlier diagnosis of dysplasia, and those screened because of known DES exposure; exclusion from the study of the 1st group reduced the frequency of all grades of dysplasia and CIS to 2.1%, and the rate of severe dysplasia to 0%. The National Collaborative Diethylstilbestrol Adenosis (DESAD) Project has also shown that rates of dysplasia were only 1.83% in DES-exposed daughters while subjects referred for screening exams had rates nearly 3/4 higher. Furthermore, 2 recent studies which compared rates of dysplasia in exposed and nonexposed cohorts showed that dysplasia was found more often in DES-exposed women, while in the DESAD project the prevalence rates were not statistically significant between exposed and nonexposed patients, but the ocurrence of dysplasia was more frequent in older women and in those who had intercourse with increased numbers of partners, a factor which was found to be important also when considering the association between oral contraceptives (OCs) and CIS. Microglandular hyperlasia can also arise in DES-exposed women, but it must not be interpreted as clear-cell adenocarcinoma, and it is associated with long-term use of OCs.

An analysis of 6055 colposcopically directed biopsy specimens from 2635 diethylstilbestrol (DES)-exposed women and 445 biopsy specimens from 277 nonexposed women was undertaken to correlate microscopic findings with colposcopic patterns. All examinations were performed using a standardized protocol which required that each participant have colposcopy, cytologic smears, and biopsy of abnormal colposcopic lesions. The findings of colposcopic "columnar epithelium, gland openings, and Nabothian cysts" correlated most often with glandular epithelium in the biopsy specimen. "White epithelium," which includes three related colposcopic patterns, mosaicism, punctation, and white epithelium, was associated most frequently (82-93% of cases) with squamous metaplasia, but occasionally with dysplasia and carcinoma in situ (CIS)(0-6%). The presence of dysplasia or CIS in any individual biopsy specimen occurred most frequently with the observation of higher graded lesions by colposcopy or a prior diagnosis of dysplasia.

A study was performed to determine if ploidy levels of nuclear DNA content could be used as a prognostic index of the biologic behavior of clear cell adenocarcinoma of the vagina and cervix in young females. Forty tumors were examined. Four patterns of nuclear DNA distribution were identified: (1) peridiploid stem cell lines (2N to 3N); (2) peritetraploid stem cell lines (3N to 5N); (3) hypertetraploid stem cell lines (greater than 5N); and (4) highly aneuploid without detectable stem cell lines. Tumors having peridiploid and peritetraploid stem cell lines (low ploidy group) were most often clinical Stage I or II (87%) and were in general associated with a favorable prognosis. Tumors having hypertetraploid stem cell lines and highly aneuploid distribution (high ploidy group) were usually clinical Stage III or IV (65%). Clinically advanced tumors (Stage III or IV) had poor prognosis irrespective of ploidy level. Among the clinical Stage I and II tumors, the low ploidy group had a slightly better prognosis than the high ploidy group. This difference, however, was not statistically significant.