Henry J. Norris

Endometrial curettings from 170 patients with all grades of endometrial hyperplasia, who did not undergo a hysterectomy for at least 1 year were evaluated in order to correlate the histopathologic features with behavior. Follow-up ranged from 1 to 26.7 years (mean, 13.4 years). Cytologic and architectural alterations were analyzed separately in order to assess their respective roles in predicting the likelihood of progression to carcinoma. Classification of proliferative lesions based solely on the presence of cytologic atypia revealed that atypia was a discriminant factor. Proliferations lacking cytologic atypia were designated hyperplasia and those displaying atypia were designated atypical hyperplasia. Only 2 (1.6%) of 122 patients with hyperplasia progressed to carcinoma compared with 11 (23%) of women with atypical hyperplasia (P = 0.001). Subclassification of the two forms of hyperplasia (those with cytologic atypia and those without) was performed using the degree of architectural abnormalities. Hyperplasia and atypical hyperplasia displaying marked glandular complexity and crowding producing a back-to-back appearance were designated complex hyperplasia (CH) and complex atypical hyperplasia (CAH), respectively. Hyperplasia and atypical hyperplasia with lesser degrees of glandular complexity and crowding were designated simple hyperplasia (SH) and simple atypical hyperplasia (SAH), respectively. Progression to carcinoma occurred in 1 (1%) of 93 patients with SH, in 1 (3%) of 29 patients with CH, in 1 (8%) of the patients with SAH, and in 10 (29%) of the patients with CAH. The differences between the four subgroups suggest a trend but are not statistically significant. The findings in this study provide a rationale for classifying noninvasive endometrial proliferations primarily on the basis of cytologic atypia since this is the most useful criterion in predicting the likelihood of progression to carcinoma. In addition, the presence of concommitant architectural alterations appears to identify a particularly high-risk subgroup.

Fifty-eight immature ovarian teratomas were studied. Neoplams with other germ cell elements (endodermal sinus tumor, choriocarcinoma, and dysgerminoma) were excluded so that the clinical and pathologic features of "pure" immature teratomas could be defined and correlated with the prognosis. The primary tumors and their metastatic growths were graded from 0 to 3. Forty were stage I; nine, stage II; and nine, stage III. The size and stage of teratomas were related to survival, but it was the grade of the primary tumor that best determined the likelihood of extraovarian spread, and it was the grade of the metastases that related best to the subsequent course. Actuarial survival was 63% at 5 years and also at 10 years. Regardless of the grade of the primary tumor, only one of six with grade 0 metastases progressed, and that neoplasms may not have been adequately sampled. Two of five neoplasms having grade 1 metastases did not progress, and two of six patients with grade 2 metastatic growths were living after relatively long intervals. All seven patients with grade 3 metastases died with tumor, none surviving more than 2.1 years. Survival of patients with grade 1, 2, and 3 neoplasms was 81, 60, and 30% respectively. The importance of adequate sampling of primary tumor and metastases for estimating prognosis and determining therapy is stressed.

The clinical and pathologic features of 71 endodermal sinus tumors of the ovary were studied in an effort to delineate the histogenesis and biologic behavior of this neoplasm and to evaluate the efficacy of different forms of treatment. Alpha-fetoprotein (AFP) was identified in hyaline droplets, cell cytoplasm, and intercellular spaces of all 15 tumors examined by an immunoperoxidase technique; this supports the view that the neoplasm simulates yolk sac endoderm. There were only nine survivors among 65 patients on whom follow-up information was available; the actuarial survival was 13% at 3 years. Of the neoplasms that recurred, 93% did so within 1 year, and of those patients who died, 93% did so within 2 years. The size and stage of the tumor had prognostic significance, but the patient's age, the mitotic activity, and histologic pattern did not. Although 71% of the patients had Stage I tumors at the time of diagnosis, subclinical metastasis was present in 84% of Stage I patients. Triple chemotherapy (vincristine, actinomycin D, and cyclophosphamide (VAC)) employed after unilateral salpingo-oophorectomy in four patients with Stage I tumors resultivors among 12 Stage I patients treated with combined surgery and radiation. The finding of AFP in all tumors in which this was evaluated suggests that serum radioimmunoassay might be useful to monitor response to therapy.

The clinical and pathologic features of 53 endometrical stromal tumors were studied to determine which pathologic characteristics were related to the clinical behavior. Morphologically, stromal tumors were divided into 2 groups: 18 tumors with pushing margins (stromal nodules) and 35 tumors with infiltrating margins (endolymphatic stromal myosis or stromal sarcoma). Stromal nodules, which were expansile, noninfiltrating lesions composed of cells similar to those found in normal endometrial stroma, were considered benign. The tumors with infiltrating margins were separated on the basis of mitotic activity. Patients with endolymphatic stromal myosis had 100% survival at 5 years and those with stromal sarcoma had 55% survival. The size of the primary tumor and presence of vein invasion showed a slight correlation with the patient's prognosis but no correlation was found with increasing degrees of cellular atypism. For patients whose disease was not controlled by hysterectomy and who had symptomatic extra-uterine tumor, x-ray irradiation appeared to be of benefit. The patients' symptoms and physical examination findings were not different from those of patients having other uterine tumors.