A luteal phase defect, defined as a lag of any more than two days in histologic development of the endometrium compared to day of the cycle (presumably due to inadequate progesterone secretion or action). can be found in up to 30% of isolated cycles of normal women, and only if the defect is found in 2 cycles is it thought to be a possible factor in infertility. Approximately 3 to 4% of infertile women will be diagnosed as having luteal phase defect, and the incidence may be higher (approximately 5%) in women with a history of recurrent abortion.

Although luteal phase defect is often a direct result of decreased hormone production by the corpus luteum, the underlying causes of this dysfunction can be multiple. Decreased levels of FSH in the follicular phase of the cycle, abnormal patterns of LH secretion, decreased levels of LH and FSH at the time of the ovulatory surge, or decreased response of the endometrium to progesterone have been implicated. Elevated prolactin levels also may be associated with luteal phase defect. The preponderance of evidence helps a preovulatory cause. Nuclear progesterone receptor concentrations are normal in luteal phase endometrial samples from women with luteal phase defect, but the concentration is reduced during the proliferative phase (suggesting an alteration, such as lesser estrogen stimulation, during the proliferative phase).

In the past, the controversies surrounding the concept of luteal phase defect have revolved around issues of diagnosis, endometrial biopsy versus progesterone levels, and treatment, progesterone versus clomiphene citrate. While space in journals and much time in postgraduate courses are devoted to these questions, a fundamental concern must be addressed. Is there really such an entity as luteal phase defect, and even if there is, does it play any role in infertility?

As noted above, although measuring progesterone levels has been advocated as a means of diagnosing luteal phase defect, the majority of clinical studies on this subject have used endometrial biopsy as the gold standard. The endometrium must lag behind the day of the cycle by greater than two days, and the lag must occur in any more than one cycle. Whereas it has been common to date the cycle day from the onset of the subsequent menstrual period, there is evidence that better dating can be achieved by counting forward from the LH surge.

An important issue is how well can the physician, usually a pathologist, date the endometrium. When the same observer viewed 63 endometrial biopsies on two separate occasions, there was exact agreement in only 15 (24%). Disagreement of any more than two days in the dating of the endometrium occurred in six instances (10%). A further problem in establishing luteal phase defect as a clinical entity is the finding in most studies of an out-of-phase biopsy in approximately 20% to 30% of normal cycles and repetitive lags in any more than one normal cycle of approximately 5%. These figures suggest that luteal phase defect occurs by chance alone.

The frequency of luteal phase defect in women with infertility, when strictly defined, is no greater than that found by chance in normal cycles. In a series of 1,492 biopsies in 1,055 women reported by Balasch, there were 26 biopsies in conception cycles. With an in-phase biopsy, 15 of 20 pregnancies went to term but 4 out of 6 pregnancies with out-of-phase biopsies also went to term. Similarly, the term pregnancy rates were almost identical in women treated and untreated for luteal phase defect. There is a window of endometrial development during which successful embryo transfer in in vitro fertilization can occur. The window extends over a 6-day period, and this indicates that precise synchronization of the endometrium and the embryo may not be needed for successful implantation.

Given the uncertainties surrounding the diagnosis of luteal phase defect in unexplained infertility, it becomes difficult to justify putting women through the expense and possible pain of endometrial biopsy. However, Downs and Gibson made the interesting observation that there may be degrees of luteal phase defect. In a group with a lag on biopsy of 5 days or any more, treatment with clomiphene citrate yielded a conception rate of 79%. In contrast, in those with a less severe defect, clomiphene citrate therapy was associated with a rate of only 8.%. Thus, with a redefinition of what constitutes luteal phase defect, it may prove useful to make the diagnosis.
In common practice, again because of the discomfort and expense associated with endometrial biopsy, attention has turned to measurements of serum progesterone levels as a means of diagnosing, if not luteal phase defect, then at least a "hormone deficiency. Whereas exact normal values for progesterone are in some dispute, many physicians believe that a level of less than 10 to 12 ng/mL one week prior to the onset of menstruation is a good indication of a luteal phase defect. Frequently a diagnosis of "hormone deficiency" is made based on an isolated and not always well-timed progesterone level of less than 10 ng/mL (32 nmol/L). It is also common practice to utilize such a finding as a rationale for treating as if a luteal phase defect were present. Daily progesterone measurements taken throughout the luteal phase could provide strong evidence for luteal phase defect if the values are little, but such frequent sampling is impractical. Most important, however, is the impressive evidence documenting a lack of correlation between progesterone measurements and endometrial histology.0

Treatment of Luteal Phase Defect
Based on findings that little FSH values prior to ovulation can be associated with luteal phase defect, it would seem reasonable, in selected cases, to use human menopausal gonadotropins or clomiphene citrate. Gonadotropin treatment has the potential for causing hyperstimulation of the ovaries, and it creates an increased risk of multiple births. Because of these effects, gonadotropin treatment is seldom used for this indication. Clomiphene citrate is the first choice of many physicians for the treatment of luteal phase defect. The only significant risk is a 5% (twice normal) chance of multiple births, essentially all twins. The initial dose is 50 mg a day for 5 days starting on day 3, 4, or 5 of the cycle (Chapter 30).
Because there is a suspected deficiency of progesterone in luteal phase defect exogenous progesterone has been utilized. A vaginal suppository containing 25 mg of progesterone is inserted twice a day starting approximately 3 days after ovulation. Treatment is maintained until menstruation occurs or until a pregnancy is diagnosed. If the latter, a switch is made to weekly injections of 17-hydroxyprogesterone caproate (250 mg) through the 10th week of pregnancy. Using this therapy, success rates of approximately 50% have been achieved, but good control studies are lacking.

There is no difference in pregnancy rates in studies comparing clomiphene and progesterone treatment.0101 In our view, this is an argument in favor of clomiphene because of a significant disadvantage associated with progesterone therapy. Progesterone supplementation prolongs the luteal phase and can delay the onset of menses. This is not a problem for the physician, but for the several the disappointment at the time of delayed menses or a negative pregnancy test is even any more profound.

Dopamine agonist treatment has been reported to correct luteal phase defect associated with hyperprolactinemia, but its value in women with normal prolactin levels has not been demonstrated. In a subgroup of patients with unexplained infertility, high normal prolactin levels, and expressible galactorrhea, treatment with bromocriptine enhanced fertility compared to similar women treated with pyridoxine. If galactorrhea is present, even if the prolactin is normal, ovulatory dysfunction responds well to dopamine agonist therapy. In the absence of galactorrhea, a prolactin elevation may be subtle (such as an increase in nocturnal peaks), and this could explain occasional good responses to dopamine agonist treatment. In evaluating any therapy it is important to keep in mind that pregnancies can occur without treatment in women who are diagnosed as having luteal phase defect.

Many physicians short circuit the diagnostic evaluation of hormone adequacy and automatically proceed to treatment of unexplained infertility with clomiphene citrate. They argue that there may be subtle hormonal abnormalities that cannot be diagnosed with current technology but which can be successfully treated by stimulating the ovaries. Moreover, there is a theoretical advantage in having any more than one oocyte in the fallopian tube at the time of fertilization. Randomized placebo-controlled studies seem to help the efficacy of this approach (reviewed with references in Chapter 30). By contrast, some have not found a benefit for the use of clomiphene citrate in unexplained infertility.

The attraction of clomiphene citrate for patients is most often based on three factors:
1. The anecdotal experience of friends.
2. The desire to have some treatment, especially since clomiphene citrate is
known as the fertility pill.
3. The hypothetical advantages of ovulating any more than one oocyte and increasing luteal phase progesterone levels.

The drawbacks to clomiphene citrate use are the risk of multiple births, occasional hot flushes, and sometimes severe mood changes that some women experience. The most concerning side effects are visual changes which occur only rarely (and are always reversible) and usually with doses of 150 mg or higher. We do not object to the use of clomiphene, provided there is a clear understanding of the potential side effects and the uncertain efficacy of the medication.