POSTCOITAL DIETHYLSTILBESTROL

In agreement with its extragovernmental physicianadvisers, FDA has approved, under restricted conditions, postcoital (contraceptive) use of diethylstilbestrol (DES), a synthetic estrogen. Adequate evidence to support the use of any other estrogen for this purpose is not presently available.

The Agency considers the use of DES for this purpose to be safe only as an emergency measure (in situations such as rape, incest, or where, in the physician's judgment, the patient's physical or mental well-being is in jeopardy) and explicitly warns against its routine or frequent use as a contraceptive.

Physicians are urged, prior to prescribing DES for this purpose, to inform patients (or guardians) fully of the possible side effects of the drug, and of alternative measures available and their hazards, so that the patient may participate in an informed way in the decision to use the drug. Pregnancy should be ruled out by appropriate tests prior to instituting therapy, so that no unnecessary exposure of a fetus to DES occurs.

The efficacy of DES in preventing pregnancy depends upon the time-lapse after coitus and dosage of the drug. The currently recommended dosage is 25 mg twice a day for 5 continuous days beginning, preferably, within 24 hours and not later than 72 hours after exposure. When this dosage is given within the specified time interval after sexual intercourse, DES is highly effective in preventing conception. But the patient must be warned to take the full course of the drug in spite of the nausea which commonly occurs, if it is to be effective.

There is at present no positive evidence that the restricted postcoital use of DES carries a significant carcinogenic risk either to the mother or fetus. However, because existing data support the possibility of delayed appearance of carcinoma in females whose mothers have been given DES later in pregnancy, and because teratogenic and other adverse effects on the fetus with the very early administration recommended are ill understood, failure of postcoital treatment with DES deserves serious consideration of voluntary termination of pregnancy.

Before prescribing, the physician should be familiar with the complete FDA-approved labeling on products intended for this use.

NITROFURAZONE LABELING REVISED TO ELIMINATE ALL-PURPOSE USE

The indications for nitrofurazone, a topically applied antibacterial sold most widely under the trade name Furacin, are being changed to eliminate all purpose use of the drug.

New indications will say that nitrofurazone is indicated in two circumstances:

(1) For adjunctive therapy of patients with second and third-degree burns when bacterial resistance to other drugs is a real or potential problem, and (2) In skin grafting where bacterial contamination may cause graft rejection and/or donor site infection, particularly in hospitals with historical resistant bacteria epidemics.

Some members of the class of chemicals to which nitrofurazone belongs are known to induce tumors of various types in rodents. Nitrofurazone itself has been shown to promote the development of mammary tumors when fed in high doses to female rats.

The relevance of these findings in man is unknown, but nitrofurazone is known to be absorbed through the skin. Topically applied nitrofurazone also carries the risk of cutaneous sensitization. There are no well-controlled studies to support the prophylactic use of topical antimicrobial agents in minor and usually self-limited cutaneous traumas.

TARDIVE DYSKINESIA ASSOCIATED WITH ANTIPSYCHOTIC DRUGS

Tardive dyskinesia, a syndrome characterized by rhythmical involuntary movements, may be associated with the use of phenothiazines and certain other antipsychotic medications. The condition appears in some patients on long-term therapy, sometimes even after it has been discontinued. FDA recommends that the antipsychotic medications be withdrawn immediately if the syndrome appears, and has added the following statement to the ADVERSE REACTIONS section of the package inserts of antipsychotic drugs:

Tardive dyskinesia may appear in some patients on long-term therapy or may occur after drug therapy with phenothiazines and related agents has been discontinued. The risk appears to be greater in elderly patients on high-dose therapy, especially females. The symptoms are persistent and in some patients appear to be irreversible. The syndrome is characterized by rhythmical involuntary movements of the tongue, face, mouth or jaw (e.g., protrusion of tongue, puffing of cheeks, puckering of mouth, chewing movements). Involuntary movements of the extremities sometimes occur. There is no known treatment for tardive dyskinesia; antiparkinsonism agents usually do not alleviate the symptoms. It is advised that all antipsychotic agents be discontinued if the above symptoms appear. If treatment is reinstituted, or dosage of the particular drug increased, or another drug substituted, the syndrome may be masked. It has been suggested that fine vermicular movements of the tongue may be an early sign of the syndrome, and that the full-blown syndrome may not develop if medication is stopped when lingual vermiculation appears.

At times a choice between two evils must be made, as when on the one hand a patient's psychosis worsens without a given antipsychotic drug whereas on the other hand administration of the drug causes severe side effects. Use of the lowest possible dose of the drug adequate to control symptoms appears to be one possible answer. Alternatively, the patient may be given another c'rug chemically least like that causing the dyskinesia.

The incidence of tardive dyskinesia is not well established. It is rarely seen in acute psychiatric units, even in patients with recurring schizophrenia. But incidence rates of about 20 percent have been reported in older, institutionalized, chronically ill patients. Perhaps 3 percent to 6 percent of patients in a mixed psychiatric population receiving neuroleptics exhibit some aspects of the syndrome at one time or another.

The physician may be able to reduce the risk of producing tardive dyskinesia by:

1. Minimizing the unnecessary use of neuroleptic medication (especially in high doses) in chronically ill patients. Many of the latter can be satisfactorily maintained for long periods without antipsychotic drugs. Drug holidays are advised in patients receving long-term medication.

2. Discontinuing neuroleptics, if possible, at the first sign of abnormal oral or lingual movements or other possible manifestations of tardive dyskinesia.

3. Observing these precautions carefully in the elderly (females especially) and probably in all patients over age 50.

REQUIREMENTS FOR PROTECTION AGAINST X RAYS

From 1300 to 6000 cancer deaths annually are caused by exposure of the American public to present levels of diagnostic x rays. In addition, ill health results from genetic damage caused by the exposure. These conclusions were reached by the National Academy of Sciences/National Research Council through its study on The Effects on Populations of Exposure to Low Levels of Ionizing Radiation. A report on the study was recently released.

Improved X-ray Technique Needed The NAS-NRC study further shows that present exposure of the popu lation to x rays, and the associated toll in lives, can be significantly reduced through simple improvements in x-ray techniques. For example, present genetic damage from x rays can be reduced by up to 50 percent through improved techniques such as using gonad shielding, particularly on male patients and restricting the size of the x-ray beam to the area of clinical interest. This is especially important

during those x-ray procedures in which the reproductive organs are in the direct x-ray beam, as during examinations of the lower back, lower abdomen, and hip. In view of the known leukemogenic effect of x rays, avoiding needless exposure of the bone marrow is also of special significance. Because of the increased radiosensitivity of embryonic tissue, special prudence should be exercised in prescribing x-ray examinations for pregnant or potentially pregnant

women,

Existing X-ray Equipment Need Not Be Upgraded FDA has asked State radiation control agencies to advise diagnostic x-ray equipment dealers and users that the new Federal radiation protection standard for x-ray machines and components does not require modification of equipment now being used. The action was taken after FDA learned that some dealers were suggesting that all existing x-ray equipment will have to be upgraded to meet the Federal standard when it becomes effective next August 15. FDA noted that, although equipment now in use will not have to be modified before the standard becomes effective, manufacturer-certified components installed in x-ray systems after August 15 must be of the type called for by the standard. All changes made in equipment and all sales must be in compliance with the regulations.

U-100 INSULIN

U-100 insulin, which is now being made available, offers distinct advantages over U-40 and U-80 preparations. It is offered as regular and in a variety of long-acting forms.

Each milliliter of U-100 insulin contains 100 units of insulin. The drug is dispensed in 10 ml (1000 unit) vials.

It is important that patients for whom this new insulin is prescribed be instructed in the use of the new 1 ml U-100 syringe. Each calibration of this syringe represents 2 units or 0.02 ml. Every 0.1 ml or 10 unit calibration is marked with corresponding numerals.

When small doses are given or accuracy down to 1 unit is necessary, a 0.35 ml (35 unit) syringe is available in which each calibration represents 0.01 ml or 1 unit.

U-40 and U-80 insulins will continue to be marketed until the use of U-100 is generally accepted.

RESPONSE TO DRUG EXPERIENCE REPORT

The FDA deeply appreciates the response to the Drug Experience Report (short form) enclosed with the last issue of the Drug Bulletin. More than 2000 adverse drug reaction reports have been received, the best short term accumulation on record. Replacement forms have been sent to all physicians who responded and a summary of the evaluated data will be sent to them as soon as it is available.

55-647 O-75-9

334

Feb. 13, 1975

IN

N 1971 a case-control epidemiologic investigation associated the recent occurrence of clear-cell adenocarcinoma of the vagina in young women with intrauterine exposure to diethylstilbestrol administered to their mothers for the therapy of high-risk pregnancies.' The results of this investigation were soon confirmed, and since then, over 170 cases of vaginal and cervical clear-cell adenocarcinomas have been collected, with a definite history of maternal exposure to non-steroidal estrogens in 65 per cent of the investigated cases. Although these,carcinomas have been encountered only rarely in females exposed to diethylstilbestrol, a variety of other abnormalities of the lower genital tract have been reported to occur with great frequency. These include vaginal adenosis (the presence of glandular epithelium in the vagina), cervical erosion, or ectropion (glandular epithelium on the portio of the cervix), and transverse fibrous ridges in the vagina and on the cervix ("cervical pseudopolyp," "cockscomb cervix," or "cervical hood").4-8 Although most of these abnormalities were known to occur rarely in the prestilbestrol era, their prevalence in unexposed women examined as carefully as those who have been exposed has not been established. The present study was undertaken to compare the results of epidemiologic, clinical, and pathological investigations of exposed women with those of unexposed controls, as well as to ascertain the current status of the exposed postmenarchal women whose mother had been treated in the Boston Lving-in Hospital Diethylstilbestrol Clinic between 1947 and 1958.

MATERIALS AND METHODS

Identification of Exposed and Control Subjects

Through the generous cooperation of Drs. Olive and George van S. Smith, accurate research records of the 841 mothers treated in the Diethylstilbestrol Clinic of the Boston Lying-in Hospital between 1947 and 1958 were made available to us. These records contain both the dates of all the pregnancies and the details of

From the Vincent Memorial Hospital (Gynecological Service of the Massachusetts General Hospital) and the James Homer Wright Pathology Laboratories, Massachusetts General Hospital (address reprint requests to Dr. Herbst at the Adenocarcinoma Registry, Warren 275, 275 Charles St., Boston, MA 02114).

Supported in part by a grant (ROI CA1 3139-03) from the National Cancer Institute and a grant (ET-52) from the American Cancer Society (Dr. Robboy is a junior faculty fellow of the American Cancer Society).

abnormality including adenosis (p < 0.0001). Abnormal cervical epithelium occurred in almost all exposed subjects but in only half the unexposed (p < 0.0001). The incidence of vaginal adenosis was highest when diethylstilbestrol was begun in early pregnancy. It was not detected when treatment was initiated in the 18th week or later. Oral contraceptive use and prior pregnancy were associated with less adenosis and erosion, respectively (p <0.05). No cases of cancer were observed. (N Engl J Med 292:334-339, 1975)

therapy. The drug was administered according to a standard schedule, with the dosage depending on the week in pregnancy in which the therapy was started. During or before the sixth week, 2.5 mg daily was given, increasing to 5 mg in the seventh week. An additional 5 mg was administered every two weeks until the 15th week, when 25 mg per day was prescribed. The daily dosage was then increased by 25 mg each month, reaching a maximum of 150 mg. All the mothers began therapy before the 23d week, and the drug was discontinued at the end of the 35th week..10 We assigned a random number to each diethylstilbestrol pregnancy, and, starting with the lowest assigned number, located the hospital birth records and then traced the parents of each female offspring. Telephone books and city directories were consulted, and the generous assistance of the Massachusetts Registry of Motor Vehicles was obtained. The correct identity of the mother was then verified by telephone or letter, and an examination of her female offspring was offered without charge if she lived within approximately 1000 km of Boston.

We selected the control group simultaneously by examining the hospital birth records and identifying an unexposed female who was born on the clinic service closest in time to (and always within five days of) each exposed female. It was further required that the mother's initial visit to the prenatal clinic occur before the 23d week of gestation. The controlled portion of the study was limited to white women 18 years of age or older, for several reasons. First of all, it appeared unreasonable to request healthy, unexposed girls under the age of 18 years to be examined. Secondly, since all but five exposed subjects were white, it was decided to limit the controlled portion of the study to whites. Finally, it was not possible within the hospital data base to match maternal gravidity and parity and still find an adequate number of unexposed subjects. Therefore, these factors were not controlled. The same technics used to find exposed subjects were used to locate controls. The results of the examination and tests with appropriate recommendations for follow-up observation were sent by letter to the subject and her physician. In addition to those included in the controlled study, all exposed post-menarchal subjects who could be located were offered appointments, to be certain that they would be examined by a physician. This precaution also permitted analysis of a larger number of cases in the second, uncontrolled portion of this investigation.

Investigation of the Subjects

Special clinics at each of which six to eight subjects were examined were conducted over an 18-month period. The two research analysts who identified and communicated with the mothers and their daughters knew which of the latter had a history of exposure to diethylstilbestrol, but this information was not available to those who performed the physical, pathological, or cytologic examinations. The maternal health histories were obtained by personal or telephone interview by the two research analysts. Each subject's medical history was obtained by the examining physician, and a thorough general physical examination was performed, followed by a pelvic examination that included

both a scrape of the vaginal mucosa and a pool aspiration for cytologic testing and careful inspection, palpation, and iodine staining of the vagina and cervix. Areas that appeared red or otherwise abnormal or that failed to stain with iodine solution were recorded and biopsied. Rectangular pieces of tissue up to 3 mm in diameter were obtained for microscopical examination with the use of either the Younge or the Kervorkian punch.

Pathological Evaluation of Tissue

Biopsies from the fornices of the vagina or areas distal to it were classified as vaginal. Those from the cervix or from fibrous ridges on or in apposition to the cervix (cervical rim or hood) were classified as cervical. At least five slides were prepared from each specimen; two were stained with hematoxylin and eosin, and one each with mucicarmine and periodic-acid-Schiff with and without diastase.

Analysis of Data

In the controlled portion of the study, the findings in the exposed and unexposed subjects were compared statistically by chisquare analysis. In the second portion of the study, the variation in the incidence of vaginal adenosis as a function of the time of initiation of diethylstilbestrol therapy during pregnancy was examined by testing for a linear trend in proportions." The findings among the exposed subjects were analyzed further by the use of a multiple logistic model12 to provide estimates of the effect of several independent variables (log of diethylstilbestrol dosage, week during pregnancy of initiation of therapy, age, history of intercourse, use of oral contraceptives, and pregnancy history of the subject) on three dependent variables (vaginal adenosis, failure of iodine staining of the vagina, and cervical erosion). This method enabled the testing of the effect of one or more independent variables after adjustment for effects of a different independent variable-e.g., the effect of use of oral contraceptives on the presence of vaginal adenosis, with adjustment for the week in pregnancy during which diethylstilbestrol therapy was begun.

RESULTS

Investigation of 110 Exposed and 82 Control Subjects 18 to 25 Years of Age

Epidemiology (subjects and parents). The 841 pregnancies of mothers treated in the Clinic yielded 731 viable births (407 male and 324 female). Among the 324 female subjects, there were 267 white women 18 years of age or older who were compared to 267 unexposed women for the control portion of the study. As shown in Table 1, aithough 192 of the exposed and 203 of the unexposed were located, only 110 exposed and 82 unexposed women agreed to be examiner. Among these, 66 per cent were 24 to 25 years and the remainder were 18 to 23 years of age. Because only those who were willing to participate in the study could be examined, it is possible that bias was introduced by the greater reluctance of the unexposed women to accept an examination. However, a comparison of the characteristics of the exposed and the unexposed subjects and those of their parents revealed a surprisingly close correspondence (Table 2). Sociologic factors such as maternal education and family social class were similar.

There was no difference in the maternal histories of serious illnesses, including cancer. Likewise, no differences were noted in the health histories of the exposed and the control daughters, none of whom had cancer; both groups were also similar in age at onset and frequency of menstruation. The reproductive histories of both groups of daughters were also similar.

Physical examination. Between the two groups of subjects several differences that did not achieve statistical significance were noted on general physical examination. Abnormalities detected included scoliosis (three exposed), Charcot-Marie-Tooth disease (one exposed), and cardiac murmurs that were interpreted as organic (nine exposed and seven control). There was a difference in the height (160.5 cm exposed vs. 158:2 cm control, p < 0.05) but not in the weight (60.0 kg vs. 56.9 kg respectively) or in the ponderal index, which integrates weight as a factor of height (12.6 vs. 12.7 respectively). Otherwise, the general physical characteristics of the exposed and control subjects were similar.

Pelvic examination. No differences were detected in the appearance of the external genitalia, or in the estimated sizes of the uterus or the ovaries. Vaginal examination, however, revealed numerous differences between the two groups that were highly significant statistically (Table 3). Transverse fibrous ridges of the vagina and cervix were seen in 22 per cent of the exposed females, but in none of the controls. These ridges usually appeared as transverse bands in the upper vagina or near the cervix. They occasionally obliterated the vaginal fornices and often obscured the boundary between the vagina and cervix. Failure of portions of the vaginal mucosa to stain with iodine solution was observed in over half (56 per cent) the exposed subjects but in only 1 per cent of the controls. Some of these abnormal areas appeared red in the exposed subjects before iodine staining, but red areas in the vagina were not observed in the controls. Also, failure of portions of the cervix to stain with iodine was noted almost twice as often in the exposed (95 per cent) as in the control subiects

Smoking

Onset of menses (12.7 vs 12.5 yr)

Frequency of menses

Use of oral contraceptives (37 vs 38%)
No, who had been pregnant (46 vs 52%)
No. who produced viable births (23 vs 28%)
History of surgery

History of cancer or serious illness

A10 05 level or less