John I. Brewer

Natural history of hydatidiform mole after primary evacuation

From 1962 to 1978, 738 patients with hydatidiform mole were referred to the John I. Brewer Trophoblastic Disease Center of Northwestern University for follow-up and human chorionic gonadotropin (hCG) testing after evacuation. There was spontaneous regression of trophoblastic disease in 596 (80.8%) of the 738 patients. Of these 596 patients, regression occurred in 11 (1.8%) by day 10 after evacuation, in 124 (20.8%) between days 11 and 30, in 255 (42.8%) between days 31 and 60, and in 206 (34.6%) between days 61 and 170. Treatment with chemotherapeutic agents was required in 142 (19.2%) of the 738 patients; 125 (16.9%) of these had invasive mole (107 nonmetastatic and 18 metastatic) and 17 (2.3%) had choriocarcinoma (13 nonmetastatic and four metastatic). All 596 patients whose hCG titers declined spontaneously to normal levels have remained well and free of disease. All 142 treated patients experienced permanent remission. Thus, all 738 patients are well and free of disease 4 to 18 years after evacuation of the hydatidiform mole. The follow-up regimen described in this report furnishes information on natural history of molar pregnancies after evacuation and provides an excellent means by which all patients can be safely managed following termination of a hydatidiform mole.

Fatal gestational choriocarcinoma. Clinicopathologic study of patients treated at a trophoblastic disease center

We studied 31 autopsied cases of gestational choriocarcinoma encountered at the Northwestern University Trophoblastic Disease Center in the past two decades to learn if the clinical and morphologic aspects of these cases have been altered by therapy. These cases were analyzed for cause of death, distribution of tumor and histologic patterns in relation to the amount of chemotherapy. Tumor hemorrhage and/or pulmonary insufficiency were the most common causes of death, irrespective of the amount of therapy although other factors including drug toxicity, sepsis, and uremia led to death in six cases. The amount of chemotherapy generally did not affect the number or distribution of metastases. Histologically, nine cases showed extensive or complete necrosis. Eighteen of the remaining tumors had typical biphasic patterns, but four patients who received multiple courses of chemotherapy had atypical patterns with a marked predominance of cytotrophoblast and infiltrative growth. These atypical patterns do not appear to be a direct result of chemotherapy but may represent a more aggressive form of this tumor. This study shows that fatal gestational choriocarcinoma can have a variety of clinicopathologic features which reflect not only the biologic capabilities of the neoplasm but also the effects of chemotherapy and prolonged disease. Cancer 50:1833‐1846, 1982.

Gestational trophoblastic disease: A comparative study of the results of therapy in patients with invasive mole and with choriocarcinoma

Abstract A study was made to compare the results of chemotherapy in patients with invasive mole and with choriocarcinoma as obtained in a Center for Trophoblastic Diseases with those obtained in 82 individual hospitals. In the Center 179 patients were treated. In the 82 hospitals 93 patients were treated, in most instances there being only one patient treated in an individual hospital. The data indicate that successful treatment is statistically better in the Center. In only one category of disease—choriocarcinoma, nonmetastatic—is the success of therapy equal in the Center and in the individual hospitals. In the 28 patients transferred to the Center, after failure of therapy in the 82 hospitals, treatment was successful in 17 (60.8 per cent). Possible reasons for the better results obtained in the Center are mentioned.

Gestational choriocarcinoma. Its origin in the placenta during seemingly normal pregnancy.

: Four new cases of primary choriocarcinoma arising in the placenta during a seemingly normal gestation were studied at the Trophoblastic Disease Center of Northwestern University. In each case the patient presented with disseminated metastases while carrying an intrauterine gestation with a normally developing fetus. All four placental primaries were small; three of the tumors were microscopic and found only after extensive sectioning. Histologically, these tumors all appeared to arise from the cytotrophoblastic cells covering the stromal portion of villi, and in some areas the involved villi retained a portion of normal investing trophoblast. This study shows that gestational choriocarcinoma unassociated with hydatidiform mole can have an early stage in which chorionic villi are present. The consistently small size of the lesions studied suggests that primary placental choriocarcinoma may frequently be overlooked or missed, and that choriocarcinoma possibly has its origin in the placenta more often than in retained or persistent trophoblast following pregnancy.

Studies of the human corpus luteum

Abstract The subject of this report is concerned with the time at which regression begins in the corpus luteum of menstruation. The life cycle of the corpus luteum of menstruation in the human being has been described and its division into four typical stages has been universally accepted. These stages are termed proliferation, vascularization, mature or blossom stage, and stage of regression (Meyer, 1911; Frank, 1914). Meyer (1911 and 1932) observed that the blossom and regression stages were imperfectly limited and stated that the beginning of regression could not be definitely recognized. Novak (1934 and 1941) stated that regression begins shortly before the onset of menstruation, about the twenty-sixth day in a patient with a twenty-eight-day ovulatory cycle. The material to be presented here indicates rather that regression begins at the termination of the so-called vascularization stage, four to six days before the onset of menstruation. In order to avoid confusion, I wish to state that I shall not describe the reported specimens in relation to days of the menstrual cycle. The variability in the length of normal cycles is so great that the dating of specimens according to this method may lead to confusion. Since that portion of the menstrual cycle between the time of ovulation and the onset of menstruation is relatively constant, I shall date all specimens by the probable days of age of the corpus luteum. Exact age of a given corpus luteum is difficult to determine. Histologic characteristics offer the best method. It is possible to evaluate the approximate age within two to three days by these histologic characteristics. For the most exact age determinations, it is necessary to study a series of specimens in this report. By this method of comparison the age of an individual specimen can be estimated more accurately. The study of the normal endometrium is also an aid in making age determinations of corpora lutea.

Early Development of Choriocarcinoma

Age. Tabulation of the ages of the patients with trophoblastic disease registered in the Albert Mathieu Chorion-epithelioma Registry is presented in Table I. The patient material is received from all parts of the United States but not all material is sent to the Registry. None has been sent on the basis of the age of the patient. In all probability this sampled data of age distribution is valid for all patients with trophoblastic disease in the United States.

Fatal gestational trophoblastic disease: An analysis of treatment failures

Forty-eight of 399 patients referred to the John I. Brewer Trophoblastic Disease Center of Northwestern University Medical School from 1962 to 1979 for treatment of gestational trophoblastic disease (invasive mole or choriocarcinoma) died. All patients who died had histologically documented metastatic choriocarcinoma. The time from pregnancy event to treatment was greater than 4 months and/or the pretreatment human chorionic gonadotropin titer was greater than 100,000 IU/L in 64% of these patients. Seventy-one percent of fatal cases developed in association with term pregnancies, abortions, or ectopic pregnancies rather than hydatidiform moles. Fifty percent of patients who died had metastases to the liver, brain, and/or peritoneal cavity when they first presented for treatment. The most common causes of death were hemorrhage from one or more metastatic sites (42%) and pulmonary insufficiency (31%). Factors primarily responsible for the treatment failures in these patients were: (1) presence of extensive disease at the time of initial treatment; (2) inadequate initial treatment; and (3) failure or presently used chemotherapy protocols in advanced disease. Secondary chemotherapy, radiation therapy to sites other than the brain, and adjuvant surgical procedures failed to improve survival in these high-risk patients.

Autoradiographic analysis of normal trophoblastic proliferation

Analysis of the cell cycle has been made possible by the introduction of high resolution autoradiographic techniques with tritiated thymidine, a specific DNA precursor. This article reports a study of the cell cycle of normal early trophoblast. Placental tissue, made available by therapeutic abortion, was maintained in organ culture. The tissue was exposed to tritiated thymidine for variable periods of time and autoradiographs were prepared by the dipping method of Messier and Leblond. By using grain counts, timing of labeled mitosis, and the appearance of label in various cells, the typical cell cycle for trophoblast was formulated. The duration of DNA synthesis was approximately 5 1/2 hours. G1, phase was 7 hours and G2 was 2 hours. The generation time was 15 hours and turnover time 3 to 4 days.

Choriocarcinoma associated with term gestation

Fifty-one patients with choriocarcinoma associated with term pregnancy were treated at the John I. Brewer Trophoblastic Disease Center of Northwestern University Medical School from 1962 through 1981. An overall remission rate of 61% was achieved: 65% for 43 patients who received all of their treatment at the center and 38% for eight patients who received treatment elsewhere before referral to the center. This remission rate was significantly less (P less than 0.005) than the 87% remission rate obtained in patients with choriocarcinoma after hydatidiform mole, abortion, or ectopic pregnancy combined. Three factors were determined which significantly influenced response to treatment in these patients: (1) time from delivery to treatment greater than 4 months (41% versus 80%, P less than 0.0005); (2) presenting symptomatology other than abnormal uterine bleeding (40% versus 87%; P less than 0.001); and (3) metastases to sites other than the lung and/or vagina (22% versus 72%, P less than 0.01). There appeared to be no advantage to treating all patients with choriocarcinoma after term pregnancy with initial multiple-agent chemotherapy unless other high-risk characteristics were present.

Observation of an organism found in patients with gestational trophoblastic disease and in patients with toxemia of pregnancy

Abstract This is an initial descriptive report of observations of multiple forms of an organism found in patients with gestational trophoblastic disease and in patients with preeclampsia-eclampsia. The worm-like forms most frequently observed have an average length of 1.0 to 1.5 mm. Larva-like forms have an average length of 150 μ; primordial eggs and egg-like forms in developmental stages range from 7 to 43 μ in diameter; and sperm-like forms are 3.5 μ or slightly smaller in size. These forms have been observed in contact smears prepared from 3 ml samples of peripheral circulating blood from both groups of patients, from trophoblastic tumor tissue, from contact smears prepared from placentas of patients with preeclampsia-eclampsia, and from umbilical cord blood of infants delivered of patients with preeclampsia-eclampsia. The various forms of this organism share morphologic characteristics of several orders of helminths, i.e., hookworms, roundworms and tapeworms. The taxonomy of these forms has not yet been determined. Until the time of taxonomic classification, the various forms will be referred to as Hydatoxi lualba . We have experimental evidence that this organism has biologic activity in BALB/c mice and in beagle dogs.