THE LUTEAL PHASE DEFECT

Transcription

1 FERTILITY AND STERILITY Copyright 1976 The American Fertility Society Vol. 27, No.4, April 1976 Printed in U.SA. THE LUTEAL PHASE DEFECT GEORGEANNA SEEGAR JONES, M.D. Department of Gynecology and Obstetrics, The Johns Hopkins University School of Medicine, Baltimore, Maryland The luteal phase defect is, by definition, a corpus luteum defective in progesterone production. As indicated by histochemical evidence, in the corpus luteum, estrogen is produced by the luteinized theca cells. The estrogen-producing function of the corpus luteum seems to be unimpaired in this disorder. The clinical manifestations of the defect include infertility and repeated first trimester abortion. The concept of the luteal phase defect as a clinical entity has been difficult fo'" gynecologists to accept. Objections to its existence as an entity revolve around four considerations. First, corpus luteum insufficiency is an unusual problem. In our infertility population, it is an etiologic factor in approximately 3.5% of the infertility cases. Although among our habitual abortion patients a luteal deficiency is identified in approximately 35% of cases, this population is highly selected for endocrine rather than anatomical problems. The second difficulty relates to the diagnostic techniques used. In our experience over a number of years, it has only been possible to diagnose a luteal defect in clinical practice, as opposed to research projects, by a properly obtained and carefully diagnosed endometrial biopsy, not by a basal body temperature chart, urinary pregnanediol, or even plasma progesterone assay. Failure to accept this diagnostic approach is probably based upon the fact that endocrinology in this country Received December 29, has formerly been in the domain ofmedical endocrinologists who have been reluctant to perform endometrial biopsies. Gynecologic pathology has been inadequately performed in many areas, adding to the difficulties in obtaining a proper reading and dating of the specimen. A third problem relates to the concept that the pathogenesis of corpus luteum defect may encompass a multitude of etiologies. Finally, a fourth difficulty derives from the specific treatment of the luteal defect. In our experience, only progesterone will act as a substitute for the physiologic progesterone in the defect and, as this must be given daily by intramuscular injection or via vaginal suppositories which are not available to most clinicians, physicians have resisted recognizing this modality of therapy and have attempted to use oral progestational agents. These agents are not only ineffective in the great majority of cases but can, in fact, be luteolytic. The lack of success in this therapeutic approach has jeopardized clinical acceptance of progesterone insufficiency as a valid basis for symptoms. The following review will consider the diagnosis, pathogenesis, and treatment of a luteal phase defect, as it has been developed in our clinic over the years. DIAGNOSIS In the first paper from our institution relating to corbus luteum insufficiency in 1949,1 it was decided that the most efficient way of diagnosing corpus luteum

2 352 JONES April 1976 insufficiency was by endometrial biopsy. This conclusion was based upon comparison of basal body temperature graphs, total urinary pregnanediol excretion, and endometrial biopsies. Throughout the years, this finding has proved valid in our hands. Although the basal body temperature can alert one's suspicion, it cannot be used as a definitive diagnostic tool. Except in the presence of a marked defect in corpus luteum function, the temperature chart may not reflect quantitative progesterone insufficiency, and vice versa. For example, some patients are notoriously poor temperature takers. Others have a poor thermogenic response to progesterone, giving the impression of a luteal defect that does not exist. Either daily urinary pregnanediol assays or daily plasma progesterone levels spanning the entire luteal phase can detect statistical differences between normal and defective luteal cycles. However, such testing is, of course, impractical for clinical purposes. The parabolic shape of progesterone excretion curves during the menstrual cycle makes it extremely difficult to use a single plasma progesterone assay or 24-hour urine pregnanediol value as an index of normalcy of luteal function. Although it has been stated that, with three serum progesterone values spotted during the luteal phase, one can statistically arrive at the diagnosis of corpus luteum insufficiency, from a practical point of view in most clinics the expense and time involved in these steroid assays are greater than for a single endometrial biopsy. The endometrial biopsy, properly obtained and properly diagnosed, can serve as a bioassay. If the biopsy is obtained on the 26th day of a 28-day cycle, almost the entire steroidogenic function of the corpus luteum is reflected in the endometrial histologic pattern. The biopsy also reflects the response of the target organ to the hormone. A biopsy should be performed using a tenaculum to steady the cervix with downward traction. The biopsy forceps should be inserted to the fundus, and a full-thickness swipe taken either anteriorly or laterally. A single strip of endometrium is usually sufficient for diagnostic purposes, and, as implantation usually occurs on the posterior surface of the uterus, if the patient happens to be pregnant at the time of the biopsy, the chance of interfering with conception is minimal. If the physician is concerned about possible interruption of pregnancy, the patient should be asked to avoid conception during the cycle to be studied. The patient must be recording basal body temperatures at the time the biopsy is performed. The biopsy should be in phase as determined by the date of the onset of the next menstrual period as well as by the estimated date of ovulation. If the biopsy is out of phase by 2 or more days, a second biopsy must be done during a subsequent cycle and the original finding confirmed. Unless this finding is consistent and repetitive, it cannot be considered as a factor responsible for infertility. Every gynecologist interested in establishing the diagnosis of a luteal phase defect should become familiar with dating the endometrium and make a habit of checking histology of the biopsies. ETIOLOGY The luteal phase defect is not caused by a single etiologic factor, but has multiple etiologies. In this respect, it resembles habitual abortion, which is also not a single disease entity. As the corpus luteum developmentally represents a continuum from the follicle, any factor affecting the follicular growth and development will also affect corpus luteum function. In our clinical experience, patients have been encountered with corpus luteum insufficiency due to (1) primary ovarian defects, (2) primary central defects of the hypothalamic-pituitary axis,

3 Vol. 27, No.4 LUTEAL PHASE DEFECT 353 (3) primary metabolic defects, and, finally, (4) specific defects in luteal cell steroidogenesis. Ovarian Factors. When we initially began to study this disease entity, we considered the etiology to be entirely central. We now recognize that a small proportion of luteal defects are specifically related to ovarian pathology. This observation was first brought to our attention by Dr. Maria Ruehsen, who, upon reviewing the histories of patients followed for as long as 20 years, found that a small percentage of women with luteal phase defects experienced premature menopause. None of the patients in this category had ever achieved a pregnancy, even with substitution therapy. Inasmuch as we consider premature menopause to represent a disturbance of oocyte maturation, we have also begun to speculate that certain patients with luteal phase insufficiency represent failure of proper follicular organization during embryogenesis, as a consequence of abnormal ova. This sequence would account for failure of pregnancy in such patients despite adequate substitution therapy.2 We have subsequently recognized several patients with chromosomal anomalies and dysfunctional bleeding interspersed with ovulatory cycles associated with inadequate luteal phase who, on ovarian biopsy, show extremely reduced numbers of follicles. 3 One such patient has also experienced premature menopause, at the age of 24 years. Central Factors. Disturbances in tonic follicle-stimulating hormone and luteinizing hormone (LH) secretion, or the periovulatory LH surge, by the hypothalamicpituitary axis can be associated with luteal phase defects. These disturbances may result from specific (1) heritable defects of the central nervous system or pituitary, (2) neurogenic scars, (3) psychogenic trauma or stress, or may be secondary to (4) nutritional or (5) chronic disease factors. As previously stated, anything interfering with normal follicular development may also interfere with corpus luteum function. Therefore, if the rise in folliclestimulating hormone level, beginning in the previous cycle and continuing during the menstrual period and into the first part of the follicular phase, is inadequate, follicular development will be inadequate, and corpus luteum function will reflect this inadequacy. Such a disturbance is probably most common in the perimenarcheal child. We first noted this phenomenon while inducing ovulation with exogenous gonadotropins in patients who had had their residual pituitary function suppressed with steroids. 4 Although it has subsequently been recognized by others, Strott et al. 5 were the first to incorporate this endocrine abnormality into a clinical entity. In our studies of ovulation induction with a follicle-stimulating hormone preparation containing only small amounts of LH, we also noted that abnormal corpus luteum function was induced following administration of inadequate amounts of human chorionic gonadotropin (HCG). This led us to believe that the adequacy of the LH surge might be an important factor in normal luteal function. Subsequently, we did in fact demonstrate this defect in several patients.6 The importance of residual tonic LH in the determination of a 14-day luteal span was first demonstrated by Van de Wiele et au while inducing ovulation in a hypopituitary patient, using LH for the ovulatory stimulus rather than HCG. Human putuitary LH has a short halflife in comparison with that of HCG. Thus, to prolong luteal function during a 14-day span, repeated injections of LH were required. It was thus concluded that residual tonic LH is an important factor in the maintenance of a 14-day luteal span. Inadequate tonic residual LH, therefore, may also account for the socalled short luteal phase. The clinical

4 354 JONES April 1976 importance of this entity has not been demonstrated. If pregnancy occurs,chorionic gonadotropin can be detected within 7 or 8 days of the ovulatory LH surge. If the corpus luteum can be rescued by HCG, a corpus luteum that functions for 7 or 8 days should be sufficient to allow for pregnancy. Metabolic Factors. In our experience, the most frequent etiologic factor in this category is inadequate blood oxygenation associated with the tetralogy of Fallot. As this defect is associated with abnormal steroidogenesis of the placenta, as well as the corpus luteum, we have postulated that the poor oxygenation of blood is responsible for inadequate steroidogenesis.s Because of this, therapy under these conditions must be continued throughout the pregnancy. This is in contradistinction to the majority of luteal defects due to other etiologies, in which therapy can be discontinued when the placental function takes over progesterone synthesis from the corpus luteum. Corpus Luteum Luteolysis. Certain progestational steroids, as well as estrogens and stilbestrol, have been shown to inhibit corpus luteum function. This inhibition is apparently related to a direct effect of the agent on corpus luteum steroidogenesis, as the rapidity of the action is such that central suppression of LH output seems improbable. Although in experimental animals it can be shown that prostaglandin, synthesized by corpus luteum cells apparently under the influence of estrogen, is also luteolytic for the luteal cells, it has been difficult to demonstrate this conclusively in the human. Jones and Wentz 9 have presented some evidence to suggest that this may be the case. Therefore, the corpus luteum may have within itself the mechanism to induce its own demise. It is tempting to postulate that patients with Halban's disease, e.g., recurrent, persistent, corpus luteum cysts, have failure in production of ovarian prostaglandin. The luteolytic action of oral progestational agents explains the inability to treat a luteal phase defect successfully with these agents. 10 THERAPY As can be deduced from the above findings, successful therapy for either infertility or habitual abortion related to luteal defects must be provided in the cycle in which pregnancy occurs. Therapy delayed until after a missed period is inadequate, as perhaps the most important function of corpus luteum secretions in reproduction is to provide for implantation. Substitution for the defective hormone, progesterone, is the treatment of choice. Progestational agents, in lieu of pure progesterone, are inappropriate for several reasons. First, many of them have been demonstrated to be luteolytic, thus interfering with endogenous progesterone production, probably at the cellular level in the corpus luteum. Second, having treated a patient for progesterone insufficiency, diagnosed on the basis of an endometrial biopsy, the adequacy of the treatment must be established by endometrial biopsy repeated during therapy to determine whether the hormonal defect has been corrected. A number of investigators have shown that progestational agents do not produce the same histologic pattern as progesterone. The adequacy of therapy thus cannot be substantiated during the use of such agents. Finally, since it is still unclear why specifically progesterone is necessary for proper implantation and pregnancy development, it is necessary to substitute with physiologic progesterone per se. Synthetic progestogens show wide individual variation in their progestational properties. Since a corpus luteum deficiency is not due to a single etiologic factor, if one wishes to correct the defect, rather than substitute for the insufficiency with progesterone, an accurate etiologic diagnosis

5 Vol. 27, No.4 LUTEAL PHASE DEFECT 355 would be required. This is always a difficult undertaking and involves many research techniques that are not ordinarily applicable to the clinical office practice. In addition, if the defect reflects underlying hypothalamic-pituitary insufficiency, administration of gonadotropic hormones is expensi ve, difficult, and hazardous. The occasional central defect can be corrected by clomiphene, which can increase pituitary gonadotropic secretion. However, in a comparative study which we conducted a number of years ago, substitutional progesterone therapy was found to be more applicable and successful than clomiphene. Nevertheless, it is important to check for correctable factors which may channel through the central pathway, such as psychogenic, neurogenic, and nutritional problems or metabolic diseases. Substitutional Progesterone Therapy. Substitutional progesterone therapy can be provided either as intramuscular progesterone in oil, 12.5 mg daily, or as progesterone suppositories,* administered either vaginally or rectally, according to the patient's preference, in a dose of 25 mg in the morning and 25 mg at night. Therapy should be begun as soon as ovulation can be diagnosed. This is usually possible by approximately the 3rd day after the onset of the temperature rise and, in a 28-day cycle, corresponds to the 16th or 17th cycle day. This amount of progesterone is sufficient to overcome the usual luteal defect and yet not sufficient to delay a normal menstrual period. If the patient has not become pregnant by the fourth cycle, a repeat biopsy is performed. If one is concerned about sacrificing treatment time, the biopsy on *Progesterone suppositories for either vaginal or rectal administration can be made by the following formulary: progesterone powder, 44 gm; polyethylene glycol 400, 2096 gm; polyethylene glycol 6000, 1392 gm. This gives a concentration of 25 mg/suppository in 1760 suppositories. Proportions must be altered according to mold size. therapy can be performed during the first therapeutic cycle. If additional progesterone is required to repair the endometrial pattern, administration must be stopped on the 14th postovulatory day, to avoid a pseudopregnancy. Human Chorionic Gonadotropin Therapy. Inasmuch as human chorionic gonadotropin is one of the most effective available luteotropic agents, it has been advocated that HCG be used therapeutically for a luteal phase defect, and indeed in some cases this is possible. Two thousand five hundred international units should be given every other day. Administration must be stopped on the calculated 12th postovulatory day, to avoid induction of pseudopregnancy. The disadvantage of therapy with chorionic gonadotropin is that, if the patient has a corpus luteum that does not respond normally to HCG, the therapy is inadequate. Patients with repeated abortions associated with luteal defects are theoretically prone to this type of corpus luteum deficiencyy SUMMARY In summary, the luteal phase defect is a deficiency of corpus luteum progesterone steroidogenesis, either in amount or duration, or both. The clinical manifestations include either primary infertility or repeated first trimester abortions. The diagnosis can only be made clinically on the basis of a well-timed endometrial biopsy that is read histologically as 2 or more days out of phase with the next period in at least two cycles. The deficiency is not due to a single etiology, and an etiologic diagnosis is often difficult or impossible to make in routine clinical practice. The treatment, therefore, usually requires substitution with progesterone. In some patients with severe defects or prolonged irregular cycles, clomiphene citrate can be used successfully and, if the patient responds to

6 356 JONES April 1976 ReG, this can also be used as a luteotropic agent. Some corpora lutea, however, cannot be "rescued" by this stimulus. REFERENCES 1. Jones GES: Some newer aspects of the management of infertility. JAMA 141:1123, Ruehsen MD, Jones GES, Burnett LS, Baramki T A: The aluteal cycle-a severe form of the luteal phase defect. Am J Obstet Gynecol 103:1059, Jones GS, Maffessoli RD, Strott CA, Ross GT, Kaplan G: The pathophysiology of reproductive failure after clomiphene-induced ovulation. Am J Obstet Gynecoll08:847, Jones GS, Ruehsen MD, Johanson AM, Raiti S, Blizzard RB: Elucidation of normal ovarian physiology by exogenous gonadotrophin stimulation following steroid pituitary suppression. Fertil Steril 20:14, Strott CA, Gargille CM, Ross GT, Lipsett MB: The short luteal phase. J Clin Endocrinol Metab 30:246, Jones GS, Madrigal-Castro V: Hormonal findings in association with abnormal corpus luteum function in the human: the luteal phase defect. Fertil Steril 21:1, Van de Wiele RL, Boguml J, Dyrenfurth I, Ferin M, Jewelewicz R, Warren M, Rizkallah J, Mikhail G: Mechanisms regulating the menstrual cycle in women. Recent Prog Horm Res 26:63, Jones GS: Luteal phase defects. In Progress in Infertility, Second Edition, Edited by SJ Behrman, RW Kistner. Boston, Little Brown & Co, 1975, p Jones GS, Wentz AC: The effect of prostaglandin F oa infusion on corpus luteum function. Am J Obstet Gynecol 114:393, Johansson EDB: Depression of progesterone levels in women treated with synthetic gestogens after ovulation. Acta Endocrinol (Kbh) 68:779, Jones GS, Aksel S, Wentz AC: Serum progesterone values in the luteal phase defects: effect of chorionic gonadotropin. Obstet Gynecol 44:26,1974