<<

MODERN TRENDS'IltEl~DS Edward Wallach, M.D. Associate Editor

FERTILITY AND STERILITY Vol. 30, No. I, July 1978 Copyright © 1978 The American Fertility Society Printed in U.S.A.

AMENORRHEA-ETIOLOGIC APPROACH TO DIAGNOSIS

PAUL G. McDONOUGH, M.D.

Division of Reproductive Endocrinology, Department of Obstetrics and Gynecology, Medical College of Georgia, Augusta, Georgia 30901

A delay in the initiation of menses at the time nated. All types of normally and ectopically lo­ of puberty or the interruption of an established cated trophoblastic proliferations should be menstrual pattern constitutes . Ces­ thought of collectively and suspected in every pa­ sation of menses normally occurs in the human tient who has a positive test for human chorionic female at the time of . Failure to in­ . If suspicion of trophoblast persists itiate menses, interruption of a normal cyclic pat­ in spite of a negative routine pregnancy test, then tern of menses, or premature cessation of menses the more sensitive ,8-subunit human chorionic constitutes unphysiologic amenorrhea. Arbitrar­ gonadotropin determination should be performed. ily classifying menarchal failure and the inter­ Excluding the presence of active, viable, pro­ ruption of a normal cyclic menstrual pattern into liferating trophoblast in all patients with primary and secondary amenorrhea is largely amenorrhea should be the first consideration be­ semantic, since the same etiologic factors may be fore other etiologies are considered. operative in either instance. Over the past decade the evaluation of amen­ The average age for first menses is 13.5 years. orrhea has been assisted by advances and refine­ Menarche at chronologie ages 11 and 15 are 2 ments in diagnostic techniques. The diagnosis of standard deviations removed from the mean.1 The ovarian failure has been facilitated by the availa­ mean for the normal ovulatory menstrual inter­ bility of sensitive serum gonadotropin measure­ val is 28 days, and ranges around that interval ments. Improved techniques of chromosomal extend from 21 to 44 days.2 The average age for banding ensure a confident diagnosis of the sex cessation of menses is 49, with ranges from 35 to chromosome abnormalities associated with 55 years. Departure from these norms constitutes amenorrhea. Laparoscopic techniques have pro­ at least temporary failure of the cycling mecha­ vided an easy means of visualizing the , nism with clinical amenorrhea. An arbitrary rule and polytomography of the sella turcica assisted of thumb as to what constitutes significant by computerized axial tomography (the CAT pathologic amenorrhea is difficult to define and scan) has improved the early diagnosis of pitui­ has very little use clinically. It is the concern or tary tumors. A specific radioimmunoassay for anxiety of the patient, regardless of the duration serum has provided a sensitive indicator of amenorrhea, that prompts her visit to the of hypothalamic-pituitary dysfunction and physician. For example, interruption of the cy­ prolactin-secreting pituitary microadenomas. cling mechanism due to pr.egnancy should be Plasma levels of testosterone, growth hormone, diagnosed as soon as possible. A delay of 3 to 6 and cortisol are readily available through com­ months in the diagnosis of other causes of mercial laboratories. The general trend is toward amenorrhea does not create any harm or undue the direct measurement of pituitary polypeptides anxiety for the patient if the presence of preg­ and steroids in blood rather than their counter­ nancy or trophoblastic disease has been elimi- parts or metabolites in urine. This general switchover from urine to blood and from milli­ Received January 3, 1978. *Reprint requests: Paul G. McDonough, M.D., Department of grams to nanograms and picograms has created a Obstetrics and Gynecology, Medical College of Georgia, 1120 new dimension that the physician must master. 15th Street, Augusta, Ga. 30901. Lastly, the isolation of gonadotropin-releasing

I 1 I I -v I 2 MCDONOUGH July 1978 hormone (GnRH) and other potent analogs pro­ TABLE 1. Etiology of Amenorrhea vides for a dynamic test of hypothalamic­ Eugonadotropic amenorrhea pituitary function. Nonfunctional In spite of these improvements in diagnostic Congenital (Rokitansky syndrome) Acquired (Asherman's syndrome) techniques, a careful history and physical exami­ Functional uterus with obstruction nation with appropriately selected laboratory Hypergonadotropic amenorrhea studies remain paramount in the evaluation of Chromosomally incompetent ovarian failure (CIOF) Chromosomally competent ovarian failure (CCOF) amenorrhea. The clinician's acumen should in­ 46,XY forms clude a perspective of the different etiologies of Swyer's syndrome (XY gonadal dysgenesis) amenorrhea, their relative frequency, and an Congenital insensitivity syndrome 46,XX forms acute awareness of any particular causes which Autosomal recessive gene may compromise the life-span or fertility of the Environmental factors patient. Autosomal abnormalities Infectious infiltrative disease Autoimmune disease Gonadotropin-resistant (Savage syndrome) ETIOLOGY 17 -Hydroxylase deficiency Hypogonadotropic amenorrhea The etiologies of amenorrhea which preclude Congenital (Kallmann's syndrome) menarche tend to be represented by genetic Acquired causes or anatomical maldevelopments of the Pituitary tumors and necrosis Drug-induced amenorrhea genital tract. In contrast, the interruption of an Psychogenic amenorrhea established cyclic pattern of menses is usually Stress-induced tonic LH (short-term) psychogenic in origin. One should appreciate the Stress-induced tonic LH (long-term) Hypogonadotropism (stress, nutrition) high frequency of genetic causes in the patient Hypogonadotropic hyperprolactinemia with primary amenorrhea and the relatively high Systemic illness-endocrine frequency of psychogenic factors in patients with Thyroid (hyper, hypo) Adrenal (hyper, hypo) secondary amenorrhea. The approach to an Systemic illness-non-endocrine evaluation of these high-frequency groups will be Amenorrhea and galactorrhea stressed. Androgen excess Adrenal tumor The spectrum of pathology seen in association Congenital adrenal hyperplasia with amenorrhea is varied, but certain well­ Virilizing ovarian tumors defined categories are recognizable. These Androgenic ovary syndrome categories can be broadly grouped into eugonad­ otropic amenorrhea, hypergonadotropic amen­ orrhea, hypogonadotropic amenorrhea, and rian anlagen. The bilateral hypoplastic discrete amenorrhea occurring in association with andro­ uteri seen in these individuals are associated with gen excess (Table 1). The separation of the total vaginal agenesis. Ovarian endocrine func­ amenorrheas into these four large groups facili­ tion and exocrine function are normal. The failure tates the diagnosis. The first two groups are more of the two Mullerian anlagen to fuse completely often associated with menarchal delay, whereas eliminates the natural stimulus for normal the hypogonadotropic group usually presents canalization of the by upgrowth of urogen­ with secondary amenorrhea. Disorders of andro­ ital sinus . Developmental abnor­ gen excess may produce primary or secondary malities of the kidneys occur frequently in these amenorrhea, but are more frequently associated individuals. The most frequent renal malforma­ with the latter. tion is a solitary ectopic kidney located in the pel­ vis. Skeletal abnormalities such as mild to severe Eugonadotropic Amenorrhea and Klippel-Feil deformity occasionally occur in association with the Rokitansky­ Nonfunctional Uterus Kuster-Hauser syndrome. The etiology of this syndrome remains obscure. The high frequency of Congenital. sporadic cases tends to incriminate an autosomal recessive gene or an environmental factor affect­ The most frequent anatomical cause of primary ing early development of the mesonephric and amenorrhea is the Rokitansky-Kuster-Hauser paramesonephric system. Identical twins discor­ syndrome. 3 • 4 This syndrome is characterized by dant for Rokitansky-Kuster-Hauser syndrome hypoplasia and failure of fusion of the two Mulle- have been reviewed by Lischke et al. 5 Discordance Vol. 30, No.1 AMENORRHEA-ETIOLOGIC APPROACH TO DIAGNOSIS 3

in monozygotic twins suggests that differential diagnosis. On a few occasions after a missed abor­ environmental factors are operative in utero. tion, nonviable hyalinized or calcific villi may Individuals with the Rokitansky-Kuster­ remain in the uterus and produce anatomical Hauser syndrome are usually asymptomatic ex­ amenorrhea. Pregnancy testing is negative in cept for menarchal delay, and present with nor­ these patients, and endometrial biopsy or curet­ mal somatic and sexual development. Total ab­ tage is necessary to establish the diagnosis. sence of the vagina or the presence of a small A form of endometrial refractoriness due to vaginal pouch is usually the only physical temporary involution of the endometrial glands finding. Cytogenetic studies confirm a normal may occur following discontinuation of oral contra­ 46,XX karyotype, and the basal body temperature ceptives or after the use of injectable synthetic graph is biphasic. The occurrence of normal ad­ steroids such as Depo-Provera. This form of renarche and normal female levels of serum tes­ amenorrhea, due to target-organ unresponsive­ tosterone exclude the diagnosis of congenital an­ ness, needs to be distinguished from the hypo­ drogen insensitivity or testicular feminizing thalamic-pituitary oversuppression syndrome pro­ syndrome. The diagnosis of Rokitansky-Kuster­ duced by oral contraceptives. Hauser syndrome can usually be made clinically, and laparoscopic visualization of the pelvic struc­ Functional Uterus with Obstruction tures is not a necessity. An intravenous pyelo­ Among patients with amenorrhea secondary to gram, however, should be performed in all pa­ tients with a presumed diagnosis of Rokitansky­ an obstructed genital tract, patients with an im­ perforate hymen or a transverse vaginal septum Kuster-Hauser syndrome because of the high in­ I,. constitute the largest group. The vast majority of cidence of associated renal malformations. Extir­ patients with an obstructed genital tract present pation of the rudimentary uterine tissue is not with primary amenorrhea; however, acquired necessary. These uteri are rarely the site of obstruction of the genital tract with secondary malignancy, uterine fibroids, covert hema­ amenorrhea may occur following cervical coniza­ tometra, or symptomatic herniation into the in­ tion or as the result of a cervical or uterine malig­ guinal canal. A vagina can be created for these I nancy. ~ patients by utilizing the Frank technique. Imperforate hymen is esaily diagnosed by the I-~ Acquired. clinical constellation of normal sexual develop­ ment, , urinary frequency, and the Patients with Asherman's syndrome also have presence of a bulging mass on the perineum with normal ovarian function as evidenced by normal the Valsalva maneuver.

I • levels of gonadal steroids and a biphasic basal Transverse vaginal septum, which represents a body temperature chart. The amenorrhea is failure of canalization of the distal third of the va­ I. caused by intrauterine adhesions that partially or gina with a proximal and hema­ completely obliterate the uterine cavity. These tometra, is not always precisely recognized.7 adhesions are usually caused by overzealous Transverse vaginal septum tends to occur in sib­ I .. postpartum curettage or induced abortions com­ lings, as evidenced by two sets of affected siblings plicated by . 6 Bleeding does not occur 4 I , in patients whom we have studied. The frequent after -progesterone treatment, and the familial occurrence of transverse vaginal septum I • diagnosis can be confirmed by hysterography or in consanguineous pedigrees suggests a genetic hysteroscopy and by obtaining characteristic fi­ etiology, probably autosomal recessive inheri­ brous tissue at curettage. The exact incidence of tance. Renal malformations may accompany a I • endometrial sclerosis, or Asherman's syndrome, transverse vaginal septum but not with the same is not known but it probably constitutes a small frequency as in the Rokitansky-Kuster-Hauser I percentage of patients with amenorrhea. syndrome. Patients with a transverse vaginal I , Tuberculosis may occasionally cause sufficient septum experience obstructive symptomatology of endometrial scarification to provide target-organ the genital tract and accompanying bladder ir­ or anatomical amenorrhea. In some instances the ritability. The pelvic pain symptoms may be mis­ I , destruction of the will produce diagnosed as appendicitis in this age group unless

I ., sclerotic changes without any significant distor­ a pelvic examination is performed. Patients with tion of the intracavitary portion of the uterus on a transverse vaginal septum have a firm pelvic hysterosalpingography. Hysteroscopy, in experi­ mass which is not always symmetric in outline. I enced hands, may be necessary for the proper Visual inspection of the introitus reveals no 1-. 4 MCDONOUGH July 1978 apparent vagina and no change in perineal con­ Hypergonadotropic Amenorrhea (Primary tour or distention of the introitus with the Val­ Ovarian Failure) salva maneuver. The latter is important in dis­ Increasing numbers of patients with amen­ tinguishing a transverse vaginal septum from an imperforate hymen. A transverse vaginal septum orrhea are being identified as having hyper­ is rigid and presents a solid core of uncanalized gonadotropic ovarian failure. The identification of tissue extending over the last 3 to 5 em of the these patients has been aided by the availability lower vagina. The proximal vagina, distended of sensitive assays for serum , im­ with trapped menstrual blood, can be felt on rec­ proved cytogenetic techniques, and the wide­ tal examination. Treatment of a transverse vagi­ spread use of the laparoscope as a technique of nal septum involves surgical dissection of the un­ gonadal visualization. The statistical importance canalized distal vagina, identification of the pat­ of this group of patients with primary ovarian ent upper vagina, and mobilization of tissue suf­ failure should be appreciated when evaluating patients with hypoestrogenic forms of amenor-· ficient to bring the upper vagina down to the level rhea. In the past, the major cause of ovar­ of the introitus. Care is necessary to prevent the ian failure has been attributed to the Turner or. newly formed vagina from assuming a disfiguring hourglass deformity that interferes with coitus. quasi-Turner phenotypes with demonstrable sex Diagnostic needling of any obstructed genital chromosome abnormalities. Radioimmunoassay tract should be performed only prior to definitive techniques for serum gonadotropins have enabled surgical correction. Preliminary needling without identification of increasing numbers of phenotyp­ definitive plans for surgery may convert a proxi­ ically and cytogenetically normal 46,XX females mal hematocolpos into a proximal pyocolpos. with amenorrhea and primary ovarian failure (Fig. 1). 8 The precise relationship between 46,XX Delays in the diagnosis of amenorrhea due to females with primary ovarian failure and those genital tract obstruction will result in continued with structural abnormalities of the sex chromo­ menstrual reflux. An aseptic inflammatory proc­ some remains to be clarified. This entire group of ess with accompanying can perma­ patients, both those with normal chromosomes or nently impair future fertility. The diagnosis of an chromosomally competent ovarian failure obstructed genital tract is one of the few (CCOF) and those with abnormal chromosomes or amenorrhea emergencies. Delays in the diagnosis chromosomally incompetent ovarian failure of Rokitansky-Kuster-Hauser syndrome, or total (CIOF), tend to be categorized as patients with vaginal agenesis, do not pose a similar threat gonadal dysgenesis or rudimentary streak since permanent is already present gonads. The implication is that a genetic etiology secondary to hypoplasia of the Mullerian system. is responsible even in chromosomally normal in­ Nevertheless, early diagnosis of total vaginal dividuals with primary ovarian failure. The latter agenesis does help to prepare the patient for the remains to be clarified. It is conceivable that development of a neovagina and provides for a phenotypically normal, chromosomally competent, frank unequivocal discussion of her sterility. 18 27 52 ABNORMAL X MOSAICS The presence of eugonadotropic amenorrhea SHORT= 51"-61" clearly emphasizes the necessity for visual inspec­ MENSES = 11.5% tion of the introitus and palpation of the . All eugonadotropic patients with biphasic basal I I body temperature charts and secondary amen­ I@ IM;.I orrhea should undergo sounding of the uterus and hysterosalpingography or hysteroscopy per­ &!LQ91 2 7 X3Y 30 NORMAL ~--~x~x~--~~ formed to rule out endometrial sclerosis or NORMAL= 63"·76" Asherman's syndrome. If hysterography is nor­ MENSES= 40% mal, an endometrial biopsy should accompany 10 20 30 40 50 TOTAL PATIENTS these studies in order to rule out instances of FIG. 1. Sex chromosome constitutions in 82 patients protracted missed abortion with residual, non­ beyond 12 years of age with primary ovarian failure. The cir­ active trophoblastic tissue. Endometrial tuber­ cled 6 represents six 45,X/46,XY mosaics with rudimentary culosis is a rare target-organ disease, but histo­ streak gonads who are nonmasculinized. The other three 45,X/46,XY individuals are masculinized with sexual am­ logic and fluorescent studies help to identify the biguity. MGD, Mixed gonadal dysgenesis. 0""¢, Sisters. (Re­ bacterium. produced with permission from reference 8.) Vol. 30, No. 1 AMENORRHEA-ETIOLOGIC APPROACH TO DIAGNOSIS 5

forms of gonadal dysgenesis represent a hetero­ principal single cell-line chromosome abnormal­ geneous group of individuals with diverse etiol­ ity in patients with ovarian failure, six 45,X/ ogies. The nexus for the clinician between the 46,XY individuals have been identified who are cytogenetically normal and abnormal forms of nonmasculinized Turner or quasi-Turner gonadal dysgenesis is gonadal failure. The mea­ phenotypes with bilateral streak gonads (Fig. 1, surement of serum gonadotropins provides the circled 6) .8 The sporadic identification of such physician with a common reference point in the nonmasculinized individuals emphasizes the need approach to this group of patients who present to search diligently for a Y cell line in all Turner with ovarian failure, either at the time of antici­ or quasi-Turner phenotypes with ovarian failure, I • pated menarche or as secondary amenorrhea later regardless of the presence or absence of mas­ in life. Patients with primary ovarian failure and culinization. This cytogenetic search should in­ amenorrhea can be placed into these two distinct clude adequate cell counts. Utilization of cur­ I • clinical categories on the basis of the cytogenetic rently available fluorescent techniques to identify findings. the Y body is important. However, this is not a

I satisfactory answer to the dilemma, since only the Ovarian Failure with Abnormal Chromosomes­ genetically inert portion of the Y chromosome Chromosomally Incompetent Ovarian Failure fluoresces intensely with quinacrine hydro­ (CIOF) chloride. Perhaps future techniques to identify Y genetic material, such as the immunologic deter­ I' A spectrum of sex chromosome karyotypes mination of H-Y antigen, may assist in the iden­ ranging from 45,X to 45,X/46,XX and 45,X/46,XY tification of patients who are at risk for dysgene­ I~ has been described in these individuals. Short tic gonadal tumors. 9 Awareness of the scope and stature is the principal clinical finding in indi­ limitations of our present techniques for identify­ viduals with primary ovarian failure and struc­ ing nonfluorescent, genetically active, Y material tural abnormalities of their sex chormosomes dictates that all chromosomally abnormal (Fig. 1). 8 Other associated somatic anomalies such amenorrheic patients be followed closely with as webbed neck or shield chest may or may not pelvic examination and periodic x-rays of the pel­ 1 .. accompany the diminished stature. These minor vis in order to detect an early dysgenetic ridge somatic anomalies and short stature may be val­ tumor in patients with an unrecognized Y cell I~ uable clues as to the presence of a sex chromo­ line. Periodic x-rays of the pelvis in such indi­ some abnormality in the prepubertal child. viduals provide for the early detection of calcifica­ Beyond the time of anticipated menarche, regard­ I" tion that indicates the presence of a gonadoblas­ less of somatic phenotype, unequivocal laboratory toma.10 I , evidence of hypergonadotropic ovarian failure is the convergence point for diagnosis. Patients with With rare exceptions, privations or deletions of I, totally absent or limited ovarian function and sex chromosome material tend to diminish stat­ structural privations of the sex chromosomes are ure regardless of the portion deleted. Most pa­ I~ collectively categorized as having gonadal dys­ tients with hypergonadotropic ovarian failure genesis. It is logically assumed that the privation and chromosomal deletions tend to be less than 63 inches tall. 8• 11 The combination of short stature I • of sex chromosomes is etiologically linked to the presence of anatomical streak gonads and func­ and primary ovarian failure seems to be causally I • tional ovarian failure. related to privations or deletions of sex chromo­ A demonstrable deletion of genetic material in some material. Normal stature and cytogenetic competence, either 46,XX or 46,XY, tend to go I~ the sex chromosome constitution still constitutes the most frequent cause of hypergonadotropic hand in hand. I. amenorrhea (Fig. 1).8 The predominant chromo­ Among patients with chromosomally incompe­ somal abnormality in this group of patients is sex tent ovarian failure who menstruate, it is not pos­ I, chromosome mosaicism. The most frequent com­ sible to draw any correlation between sex chro­ binations have been 45,X/46,X,i(Xq) and 45,X/ mosome morphology and primary or secondary 46,XY. The latter group is important because of amenorrhea. This reflects the lack of an aU-or­ I • its relative frequency among mosaic patients with none phenomenon in those factors affecting ovar­ ovarian failure, and identification of the Y ian hypoplasia. 12 This broad spectrum of ovarian I" chromosome in a phenotypical female requires function probably represents varying degrees of prophylactic extirpation of the rudimentary interference with germ-cell migration, mitotic ac­ I streak gonads. Although 45,X still constitutes the tivity in the genital ridge, abnormal meiotic pair-

I .. 6 MCDONOUGH July 1978 ing, or overutilization of primary oocytes. The de­ stay open in the absence of ovarian steroids and gree of interference with each of these events the presence of a normal genetic complement. The 1 cannot be specifically related quantitatively to normal chromosomal karyotype ensures normal the deletion of sex chromosome material. stature which is further augmented by delayed In addition '

carotid artery aneurysms, and primary prognosis with spontaneous cure. In "postpill .. hypothyroidism. Secondary enlargement of the amenorrhea" due to hypothalamic-pituitary over­ pituitary may develop in primary hypothyroid­ suppression, most patients spontaneously recover ism, presumably as a compensatory mecha­ menstrual function within several months. In nism. The measurement of thyroid-stimulating others, ovulation and consequent pregnancy can hormone may help to differentiate sellar en­ be induced by ovulation-inducing drugs. In some largement due to primary hypothyroidism from instances "postpill amenorrhea" is refractory to a true intrasellar tumor. The former condition re­ all forms of ovulation induction except for human sponds readily to thyroid replacement therapy.25 menopausal gonadotropin. Naturally, it is essen­ Postpartum ischemic infarction and necrosis of tial to rule out other serious endocrinopathies, the pituitary can occur in association with obstet­ e.g., pituitary neoplasia, prior to assuming that ric shock syndromes (Sheehan's syndrome). This amenorrhea is causally related to prior pill use. form of hypogonadotropic amenorrhea is usually associated with varying degrees of insufficiency of Psychogenic Amenorrhea. other pituitary tropic hormones and is manifested Psychogenic stress factors have been recognized clinically by lactation failure, genital , as capable of producing amenorrhea. Nevertheless, hair loss, hypotension, hypoglycemia, and the precise biochemical intermediaries involved anemia. Undetectable serum gonadotropin levels in psychic alterations and gonadotropin produc­ and low T4 , thyroid-stimulating hormone, and tion are unknown. The anatomical areas within plasma cortisol levels corroborate the clinical the limbic system and hypothalamus that affect diagnosis of hypopituitarism due to Sheehan's emotional expression are identical with those • syndrome. which contain steroid receptors and some en­ zymes involved in steroid metabolism. A hypoth­ Drug-Induced Amenorrhea. esis based upon neuroendocrine control. Various drugs can induce amenorrhea by virtue mechanisms suggests that psychic stress may of their effect on the hypothalamus or through produce a chronic dopaminergic effect with in­ creased neurotransmission and a predominance weight loss secondary to anorexia. Phenothiazine of dopamine over norepinephrine. derivatives, reserpine, and ganglion-blocking agents affect the hypothalamus and produce Psychogenic amenorrhea may occur in associa­ amenorrhea which is sometimes associated with tion with either acute or chronic emotional stress. galactorrhea. These effects are usually reversible Psychogenic amenorrhea may also have a strong once the drug is discontinued. Other drugs such nutritional component related either to weight as alcohol, digitalis preparations, cytotoxic drugs, loss or weight gain. Neuroendocrine aberrations and certain antibiotics may produce amenorrhea produced by stress may produce temporary or pro­ through their anorectic effect. longed disturbances in gonadotropin and prolac­ The second category of drug-related amen­ tin production. These disturbances producing orrhea encompasses oversuppression syndromes psychogenic amenorrhea tend to fall into four following the use of oral or parenteral synthetic basic patterns: steroids. Less than 1% of women taking oral con­ Stress-Induced Tonic LH (Short-Term). Tem­ traceptives develop amenorrhea during or after porary interruption of the recycling mechanism the use of these agents regardless of the duration may occur as the result of sudden psychic stress. of their use. In these women the hypothalamic­ Cyclic LH is replaced by stress-induced, continu­ pituitary axis is suppressed by the exogenous ous LH production. In most instances failure of steroids, and inhibition may persist after the the recycling mechanism is brief in duration and medication is discontinued. This condition of there is prompt recovery after 1 or 2 months. oversuppression may persist for prolonged periods Stress-Induced Tonic LH (Long-Term). If stress of time. It is important to distinguish amenorrhea factors persist, long-term tonic LH production secondary to pituitary suppression from "postpill will produce a clinical picture not unlike polycys­ amenorrhea" caused by endometrial involution. tic . Amenorrhea will persist and In the latter condition, endogenous estrogen the patient may note mild facial hirsutism with levels are normal but the endometrium is unre­ . Continued pulsatile LH production will in­ sponsive, possibly owing to the glandular regres­ crease ovarian production of Ll4-androstenedione, sion induced by the synthetic steroid. Postpill testosterone, and estrogen. Vaginal hormonal target-organ amenorrhea usually has a good cytology will reveal many superficial cells, and 10 MCDONOUGH July 1978 bleeding will follow administration of progesta­ explain some of the aspects of the anorexia syn­ tional agents. Hyperprolactinemia has been de­ drome.27 scribed in some patients with persistent LH Even in apparent "postpill amenorrhea," the elevations, suggesting a condition not unlike role of weight loss and psychologic factors should pseudocyesis. receive serious consideration. Women oflow body Stress-Induced Hypogonadotropism. Eventu­ weight, for example, may be at particular risk for ally, psychic stress may induce loss of both cyclic developing "postpill amenorrhea." LH and tonic LH release from the pituitary. This Stress-Induced Hypogonadotropic Hyperprolac­ condition may result in low base line levels of tinemia. Some individuals with stress-acquired FSH and LH. Acquired hypogonadotropism of this hypogonadotropism have elevated levels of serum type also may occur as the result of nutritional prolactin. The hyperprolactinemia may precede factors, especially weight loss with or without or follow the hypogonadotropism. Pituitary significant psychopathology. tumors should be the first consideration in hyper­ Anorexia may be a symptom of any systemic prolactinemic patients, but stress-related neuro­ disease or occur as the result of the use of drugs or endocrine dysfunction may produce a comparable 28 medications. The weight loss which occurs secon­ pituitary profile. Time and further study should dary to anorexia may produce amenorrhea with help to delineate the true frequency of stress­ low gonadotropin levels. Anorexia in the absence induced hyperprolactinemia with amenorrhea. of illness or drug therapy is usually psychogenic Psychogenic amenorrhea may be associated in origin. Psychogenic anorexia with weight loss with elevated LH levels and low FSH, low LH and is usually seen before 25 years of age. These indi­ normal or slightly elevated FSH, or low viduals have lost at least 25% of their original gonadotropin/high prolactin profiles. These pat­ body weight and exhibit a distorted attitude to­ terns are usually seen in different patients, but a ward eating and weight gain that over-rides all single individual may pass sequentially through hunger. They take pleasure in refusing food and any combination of these different patterns. losing weight. Such individuals are hyperkinetic Systemic Illness-Endocrine (Excessive or In­ with overactivity and have an inapppropriate adequate Adrenal or Thyroid Function). perception of body image. Physical examination Amenorrhea may occur as a result of a systemic may reveal light lanugo hair growth, especially endocrinopathy involving an extragonadal endo­ over the back and malar eminences. Bradycardia crine gland. Abnormalities in thyroid function, and low body temperature are present. either hyperthyroidism or hypothyroidism, may Gonadotropin profiles in these patients with produce amenorrhea. Adrenal cortisol overpro­ weight-loss amenorrhea reveal low LH values duction and underproduction may also interfere and low or normal FSH levels in serum. There is a with normal cyclic gonadotropin production and failure of pulsatile LH output, especially during produce amenorrhea. Thyroid evaluation is an es­ sleep, and a delayed LH response to gonado­ sential part of the diagnostic work-up of tropin-releasing hormone stimulation. The ACTH amenorrhea, whereas plasma cortisol values are adrenal axis is altered similarly by decreased pul­ indicated when the clinical picture-including satile ACTH and some loss of diurnal variation in blood pressure and blood sugar-suggest cortisol plasma cortisol levels. The metabolic clearance overproduction. rate for cortisol is reduced, apparently in an at­ tempt to maintain internal homeostasis. Elevated Systemic Illness-Non-Endocrine. 3erum levels of growth hormone are seen in some The precise role of systemic illnesss in the etiol­ of the patients with this type of anorexia. Other ogy of amenorrhea is not always easy to under­ findings which have been described include stand. Generalized systemic illnesses involving hypercholesterolemia, carotenemia, hypogly­ poor nutrition, malabsorption syndromes, cardiac cemia, low serum T 3 levels, and increased disease, renal disease, severe , or neo­ catechol estrogens. 26· 27 The increased formation plasia may produce amenorrhea. It is difficult to of catechol estrogens in these patients is impor­ delineate whether the amenorrhea is directly re­ tant since they may compete centrally for estro­ lated to the illness or is the product of psychogenic gen binding sites and inhibit the biologic inacti­ factors, weight loss, or associated drug therapy. In vation of catecholamines. Prolongation of most systemic disease, anorexia may be largely catecholamine activity could provide for con­ responsible for the hypogonadotropism. The en­ tinued stimulation of dopamine receptors and docrine abnormalities seen in anorexia nervosa Vol. 30, No.1 AMENORRHEA-ETIOLOGIC APPROACH TO DIAGNOSIS 11

usually cannot be distinguished from those en­ tients with androgen overproduction involves countered in starvation due to other causes. principally the measurement of urinary 17- Amenorrhea should not be attributed to any sys­ ketosteroids, serum testosterone, and serum LH. 29 I~ temic illness or psychic disorder until other In a small number of patients with elevated 17- causes have been eliminated. ketosteroid levels, determinations of urinary pregnanetriol values or plasma 17 -hydroxy­ Amenorrhea and Galactorrhea. I~ progesterone are indicated to rule out "late onset" Amenorrhea and galactorrhea are cardinal or "late diagnosed" congenital adrenal symptoms in several syndromes-Chiari­ hyperplasia. 4 Frommel syndrome, Ahumada-Dei Castillo syn­ drome, and Forbes-Albright syndrome. Although these eponyms have been in common usage, they DIAGNOSTIC APPROACH currently serve very little purpose, and these dis­ General orders can be collectively categorized as amen­ I ~ orrhea-galactorrhea syndromes. The development History of a sensitive, specific radioimmunoassay for From the previous discussion, it is apparent serum prolactin has aided in the identification of that historic factors are paramount in evaluating those patients with amenorrhea-galactorrhea the patient with amenorrhea. Histories of drug syndromes who may have prolactin-producing ingestion, including oral contraceptives, and pituitary tumors. All patients with elevation in psychic stress must receive primary considera­ serum prolactin, regardless of the presence or ab­ tion. Careful questioning should be directed to­ sence of galactorrhea, should be identified and ward marked changes in body weight. The evaluated for a possible pituitary tumor. psychosocial history should focus on the time span I~ Androgen Excess immediately preceding the development of amen­ orrhea. The last category of patients with amenorrhea Physical Examination I~ includes those with disorders of androgen over­ production, either adrenal or ovarian. Adrenal an­ Most patients with amenorrhea have a normal drogen overproduction may occur as a result of a physical appearance. The departures from normal virilizing adrenal tumor or congenital adrenal that should receive careful attention include hyperplasia. Virilizing ovarian tumors are rare. signs of (1) androgen overproduction, (2) cortisol The principal cause of ovarian androgen over­ overproduction, (3) galactorrhea, (4) hypo­ production is the androgenic ovary syndrome, or thyroidism, (5) acromegaly, (6) weight loss and/or I~ polycystic ovaries, an etiologically complex and weight gain, (7) short stature-with or without clinically heterogeneous condition. Patients with associated somatic anomalies, and (8) sexual in­ the polycystic ovary syndrome usually have nor­ fantilism. The last three observations include a mal thelarche. Pubarche is followed by sequential careful initial recording of the patient's height, I~ growth of hair on the upper lip, chin, intermam­ weight, and degree of sexual development. A care­ mary area, and extremities. Serum LH levels are ful search is made for any somatic anomalies tonically elevated with normal or slight elevations (such as webbed neck) that might additionally I~ of serum testosterone and urinary 17 -ketosteroids. suggest ovarian failure associated with a sex The vaginal smears are remarkably estrogenic, chromosome deletion. with 25% to 40% superficial cells. I~ Presumably, this group of hirsute patients with Evaluation of Endogenous Estrogen amenorrhea has a constant tonic "leak" of LH The pivotal point in the diagnosis of all which stimulates ovarian production of the weak amenorrhea is an evaluation of the patient's en­ androgen a 4-androstenedione and produces mor­ dogenous estrogen production. This can be ac­ phologic polycystic ovaries. Increased production complished clinically by utilizing examinations of rates for testosterone and follow as a re­ the vaginal smear, cervical mucus, or endome­ sult of the increasing formation of the prehor­ trium to bioassay the patient's estrogen capacity. mone a 4-androstenedione. Increased testosterone Of these three bioassays, the vaginal smear is I~ production results in hirsutism, while an in­ probably the most sensitive index, since it creased production rate for estrone provides for a exhibits a relatively linear response to increases well-estrogenized genital tract. Evaluation of pa- in endogenous estrogen. Estrogenic quantitation 12 MCDONOUGH July 1978 by endometrial histology is difficult, and endocer­ appropriate feedback hormone, namely FSH. In vical mucus may be limited by the frequent oc­ unusual patients the target-organ effects of estro­ currence of endocervicitis. A progesterone chal­ gen may be normal but serum FSH levels are lenge test is the most frequently used dynamic markedly elevated. These unusual patients are test of endogenous estrogen. An appreciation of categorized as having euestrogenic forms of the scope and limitations of the progesterone ovarian failure.30 They serve to emphasize the dif­ challenge is important. The progesterone chal­ ferential sensitivity to estrogen of target tissues, lenge must be performed with a "pure or C-21" including the hypothalamus. Differential target­ progestin such as progesterone in oil or medroxy­ organ sensitivity and limitations of the estrogen progesterone acetate. The 19-norsteroid com­ assay point out the need to measure serum FSH pounds are not reagent-pure and may have some in clinically suspected . intrinsic estrogen contamination, either in their preparation or metabolism. The response to 10 mg Specific of medroxyprogesterone acetate daily for 5 days The evaluation of any patient will depend upon or to 100 mg of progesterone in oil is a suitable the assessment of endogenous estrogen. On the test, but the response must be carefully inter­ basis of the vaginal smear or progesterone chal­ preted to be certain that the withdrawal menses lenge, patients with amenorrhea can be separated is of sufficient quantity and time span to suggest prognostically into those with normal and low en­ adequate endogenous estrogen levels. Patients dogenous estrogen levels. The prognosis for spon­ with relatively low levels of endogenous estrogen taneous cure is good in the normal group and may occasionally manifest a positive but limited guarded in the hypoestrogenic category. Any diag­ response to a progesterone challenge. Conclusions nostic work-up for amenorrhea should be prefaced about endogenous estrogen on the basis of proges­ by a concern for the "target-organ pathology" of terone withdrawal must be guarded in limited re­ pregnancy. A ,8-subunit assay for human sponse situations. Target-organ failures are best A chorionic gonadotropin should be obtained if evaluated by prolonged high-dose estrogen chal­ there is any suspicion of active, viable, tropho­ lenges. Ultimately, the acquired form of endome­ blastic tissue. trial sclerosis must be diagnosed by hyster­ The diagnostic work-up of amenorrhea, ography or hysteroscopy. Failure to bleed follow­ categorized on the basis of endogenous estrogen ing estrogen priming does not always indicate the levels, is outlined in Table 3. The sequential presence of target-organ disease. Long-term work-up of each group is designed to uncover the hypoestrogenic patients may have a refractory most frequent and serious etiologies associated endometrium which requires substantial with normal or low estrogen production. Etiologic amounts of estrogen priming before displaying a overlap occurs frequently, since temporary eues­ positive response. This is especially true in the trogenism is a transitional stage for severe forms weight-loss hypoestrogenic amenorrheas. Over­ of hyper- and hypogonadotropism. production of catechol estrogens in these patients may have an anti-uterotropic effectY Low-Estrogen Group The measurement of peripheral estrogenic Patients with low endogenous estrogen levels steroids has a very small practical role in the must be evaluated for central nervous system dis­ quantitation of endogenous estrogen. The inter­ ease or for hypergonadotropic forms of ovarian mediary metabolism of estrogens is extremely failure. The initial evaluation includes skull complex. The interpretation of a single measure­ x-rays and measurement of serum levels ofT4, T3 , ment of the blood level of a specific estrogen is and gonadotropins. Plasma cortisol levels are ob­ difficult and of limited value in patient manage­ tained if the clinical phenotype suggests cortisol ment. Normal or low estrogen values are difficult overproduction or if the blood pressure or blood to interpret, but unusually high serum values for sugar is elevated. The measurement of serum total estrogen or estradiol may be helpful in gonadotropins, specifically FSH, is necessary in suggesting an estrogen-producing ovarian tumor. all hypoestrogenic patients. If serum gonadotro­ The clinician is best served by using the bioassay pin levels indicate hypergonadotropism, then principle as exemplified by the vaginal smear or cytogenetic studies should be performed on all pa­ the progesterone challenge test. In certain situa­ tients, especially short-statured individuals (be­ tions, indirect quantitation of peripheral estrogen low 63 inches in height).4 If serum gonadotropin can be obtained by measuring serum levels of the levels are low, one should pause temporarily in Vol. 30, No.1 AMENORRHEA-ETIOLOGIC APPROACH TO DIAGNOSIS 13

TABLE 3. Schematic Outline of the Diagnostic Evaluation of Amenorrhea Initial step: rule out presence of trophoblast (J3·subunit human chorionic gonadotropin)

Low Normal Consider principally central nervous system Consider principally temporary failure of recycling Skull: lateral, posteroanterior films Skull: lateral, posteroanterior films T4 , T3 , blood sugar, plasma cortisol determinations T4, T3 , blood sugar, plasma cortisol determinations FSHILH Before proceeding further, give alternate-month therapy; then High: cytogenetic studies Serum LH/testosterone determinations Low: intensive psychosocial evaluation 17 -Ketosteroid determinations Before proceeding further, expand psychosocial history; then Endometrial biopsy if vaginal smear reveals 40% or more Serum prolactin, 17 -ketosteroid determinations superficial cells Growth hormone determinations, tomograms, visual field examination, CAT scan the diagnostic work-up and intensify the psycho­ drogens. A few patients in the normal-estrogen social-nutritional history. The frequency of group may have acquired target-organ disease psychogenic amenorrhea warrants this expense of due to endometrial sclerosis. The evaluation of time before embarking on further sophisticated the euestrogenic amenorrheic patient is provided endocrine and central nervous system evaluation. in Table 3. Although endogenous estrogen levels The majority of patients with acquired mono­ in this group are normal, screening of the central tropic failure of gonadotropins will be found to nervous system should be performed in order to have psychogenic amenorrhea or amenorrhea re­ eliminate patients who may have early pituitary lated to weight loss. If hypogonadotropism per­ tumors. Evaluation for hypothyroidism is still sists, then serum prolactin levels must be ob­ necessary since hypothyroid patients may be tained to identify the "silent hyperprolactinemic" euestrogenic. individuals who are suspect for prolactin-pro­ Plasma cortisol levels are obtained in the ducing pituitary tumors. The measurement of presence of clinical evidence of hypercortisolism. 17-ketosteroids in urine may be indicated in rare Interference with ovarian estrogen production patients with hypogonadotropism in order to may be a late finding in some disorders with cor­ identify androgen overproduction in the absence tisol excess. If skull x-rays and thyroid evaluation of hirsutism. The latter may occasionally be im­ are normal, a pause in the evaluation is justified portant in the diagnosis of rare adrenal tumors while a pure progestin or C-21 compound is ad­ when clinical evidence of androgen overproduc­ ministered on a cyclic basis. A synthetic progestin tion is limited. If hypogonadotropic hyperprolac­ such as medroxyprogesterone acetate, 10 mg tinemia persists, then growth hormone levels at daily, can be administered for the first 5 days rest and after 30 minutes of exercise should be every other month over a 6-month time span. One obtained. Further endocrine evaluation should be should be reasonably certain from the appearance combined with visual field studies and poly­ of the vaginal smear that withdrawal bleeding will tomography of the sella turcica. The combination occur. Failure to bleed after the medication may of tomographic abnormalities, persistently low generate further anxiety and concern on the part gonadotropin levels, elevated serum prolactin of the patient, if not adequately prepared. With­ levels, and blunted growth hormone responses to drawal menses following the administration of provocation testing suggests a central nervous sys­ medroxyprogesterone acetate in alternate months tem lesion. Decisions for further invasive studies serves to relieve anxiety, confirms that endogen­ of the central nervous system are individualized ous estrogen is adequate, and indicates a patent, on the basis of this information and the clinical responsive genital tract. During the alternate course of the patient's illness. months in which medroxyprogesterone acetate is not given, spontaneous menses may occur. The Normal-Estrogen Group vast majority of patients who have temporary in­ The vast majority of patients with normal en­ terruption of the recycling mechanism will usu­ dogenous estrogen is composed of patients who ally experience spontaneous menses during or fol­ experience temporary interruptions of the recy­ lowing the alternating 6-month trial of a pure cling mechanism with tonic elevation of LH. In progestin. some instances the tonic LH elevation may per­ If there is clinical evidence of androgen over­ sist with resultant ovarian overproduction of an- production, then serum LH, serum testosterone, 1 14 MCDONOUGH July 1978

and urinary 17-ketosteroids should be studied. In 6. Asherman JG: Amenorrhea traumatica (atretica). J some patients with androgen overproduction, the Obstet Gynaecol Br Emp 55:23, 1948 7. Akinkugbe A: Vaginal atresia and . .. vaginal smear exhibits more than 4~ superficial Obstet Gynecol46:317, 1975 cells. Endometrial biopsy should be performed to 8. McDonough PG, Byrd JR, Tho PT: Phenotypic and 1 rule out undue degrees of endometrial hyper­ cytogenetic findings in 82 patients with primary ovarian plasia. Endometrial biopsy and uterine sounding failure-changing trends. Fertil Steril 28:638, 1977 are also indicated in those patients who have 9. Saenger P, Levine LS, Wachtel SC, Korth-Schutz S, adequate estrogen effect as judged by the vaginal Doberne Y, Koss GC, Lavengood RW Jr, German JL III, New MI: Presence ofH-Y antigen and testis in 46,XX true smear but who do not bleed after a progesterone hermaphroditism, evidence for Y chromosomal function. J challenge. These include patients who may have Clin Endocrinol Metab 43:1234, 1976 acquired forms of endometrial sclerosis (Asher­ 10. McDonough PG, Greenblatt RB, Byrd JR, Hastings EV: man's syndrome) or a missed abortion with re­ Gonadoblastoma (gonocytoma III): report of a case. Obstet sidual hyalinized villi. Gynecol 29:43, 1967 11. McDonough PG: Primary ovarian failure. In Endocrine Causes of Menstrual Disorders: Proceedings of the Second SUMMARY Annual Symposium on Gynecologic Endocrinology, Edited by J Givens. Chicago, Year Book Medical Pub­ Amenorrhea is a ubiquitous problem, and lishers, 1978 clearly tangible causes are evident only in a rela­ 12. McDonough PG, Byrd JR: Gonadal dysgenesis. Clin Obstet Gynecol 20:565, 1977 tively small number of patients. The clinician 13. Espiner EA, Veale AM, Sands VE, Fitzgerald PH: Fa­ should proceed cautiously and select appropriate milial syndrome of streak gonads and- normal male laboratory studies which will be of maximal bene­ karyotypes in five phenotypic females. N Engl J Med fit to the patient. While the evaluation of en­ 283:6, 1970 dogenous estrogen, skull x-rays, and serum 14. Swyer GIM: Male pseudohermaphroditism: a hitherto undescribed form. Br Med J 2:709, 1955 gonadotropin levels are in progress, a continued 15. Greenblatt RB, Byrd JR, McDonough PG, Mahesh VB: dialogue with the patient must continue in order The spectrum of gonadal dysgenesis. Am J Obstet Gynecol A to identify factors that may contribute to 47:351, 1967 psychogenic amenorrhea. Continued studies in 16. McDonough PG, Tho PT, Byrd JR: Twins discordant for the area of neuroendocrinology may help to 46,XX gonadal dysgenesis. Fertil Steril 28:251, 1977 17. Cohen MM, Capraro VJ: Pericentric inversion of a group clarify the relationship between the functions of D chromosome (13-15) associated with amenorrhea and the neocortex and gonadotropin production. Ad­ gonadal dysgenesis. Am Hum Genet 30:313, 1967 vancements in this area should help the clinician 18. Kleerekoper M, Bastan H, Penny R, Posen S: Idiopathic in his attempts to separate dysfunction from or­ hypoparathyroidism with primary ovarian failure. Arch ganic pathology. Meanwhile, the approach to Intern Med 134:944, 1974 19. Jones GS, Moraes-Rhuehsen MD: A new syndrome of amenorrhea should be tempered by a constant amenorrhea in association with hypergonadotropism and vigilance for pituitary tumors. The physician apparently normal ovarian follicular apparatus. Am J must always be aware of the role of psychosocial Obstet Gynecol104:597, 1969 and nutritional factors in the interruption of the 20. Koninckx PR, Brosens lA: The "gonadotropin-resistant cyclic mechanism. ovary" syndrome as a cause of secondary amenorrhea and infertility. Fertil Steril 28:926, 1977 21. DeLange WE, Weeke A, Artz W, Jansen W, Dorenbos H: REFERENCES Primary amenorrhea with hypertension due to 17- hydroxylase deficiency. Acta Med Scand 193:565, 1973 1. Marshall WA, Turner JM: Variations in pattern of puber­ 22. McDonough PG, Mahesh VB, Byrd JR: The diagnostic use tal changes in girls. Arch Dis Child 44:291, 1969 of gonadotropins. Fertil Steril 2:126, 1970 2. Ross GT, Cargille CM, Lipsett MB, Rayford PD, Marshall 23. Boyar RM, Wu RHK, Kapen S, Hellman L, Weitzman ED, JR, Strott CA, Rodbard D: Pituitary and gonadal hor­ Finkelstein JW: Clinical and laboratory heterogeneity in mones in women during spontaneous and induced ovula­ idiopathic hypogonadotropic . J Clin En­ tory cycles. Recent Prog Horm Res 26:1, 1970 docrinol Metab 43:1268, 1976 3. Griffin JE, Edwards C, Madden JD, Haxzod MJ, Wilson 24. Kulin HE, Santner SJ: Timed urinary gonadotropin mea­ JD: Congenital absence of the vagina. The Mayer­ surements in normal infants, children and in patients Rokitansky-Kuster-Hauser syndrome. Ann Intern Med with disorders of sexual maturation. J Pediatr 90:760, 85:224, 1976 1977 4. McDonough PG: Menarchal delay. In Current Problems in 25. Keye WR, Ho Yuen B, KnopfRF, Jaffee RB: Amenorrhea, Obstetrics and Gynecology, Edited by RW Kistner. hyperprolactinemia and pituitary enlargement secondary Chicago, Year Book Medical Publishers, 1977, p 1 to primary hypothyroidism. Obstet Gynecol 48:697, 1976 5. Lischke JD, Curtis CH, Lamb EJ: Discordance of vaginal 26. Warren MP, Vande Wiele RL: Clinical and metabolic fea­ agenesis in monozygotic twins. Obstet Gynecol 41:920, tures of anorexia nervosa. Am J Obstet Gynecol 117:435, 1973 1973 Vol. 30, No.1 AMENORRHEA-ETIOLOGIC APPROACH TO DIAGNOSIS 15

27. Fishman J, Boyar RM, Hellman L: Influence of body 29. Yen SCC, Vela P: Inappropriate secretion ofFSH and LH weight on estradiol metabolism in young women. J Clin in polycystic ovarian disease. J Clin Endocrinol Metab Endocrinol Metab 41:989, 1975 30:435, 1970 28. Zacur H, Chapanis WP, Lake CR, Ziegler M, Tyson JE: 30. Falk RJ: Euestrogenic ovarian failure. Fertil Steril Galactorrhea-amenorrhea: psychological interaction with 28:502, 1977 neuroendocrine function. Am J Obstet Gynecol 125:850, 1976

..