Introduction

PCOS has had a complex history punctuated by a constant improvement in its understanding and evolution of diagnostic criteria . These include an awareness of forms of PCOS that are secondary to other endocrinopathies, and therefore the need for them to be recognized in the diagnostic approach to PCOS and, when possible, treated with tailored therapies . The management of secondary forms of PCOS is now possible and also thanks to the availability of increasingly precise, and extensive laboratory support . In this chapter, we describe the well-defined endocrinopathies that lead to secondary forms of PCOS, in particular, hyperprolactinemia, thyroid disorders, nonclassic congenital adrenal hyperplasia due to 21-hydroxylase deficiency (21-NCAH), Cushing’s syndrome, acromegaly, and severe insulin resistance states together with the recommended diagnostic approach to differentiate between these forms and PCOS.

PCOS—Secondary to hyperprolactinemia

Although there are no exhaustive studies investigating the association between PCOS and hyperprolactinemia, the literature available describes a high prevalence (50%–67%) of hyperprolactinemia in women with clinical, hormonal, or ultrasound features of PCOS. Prolactin exerts an important effect on the ovary, and physiologically, it stimulates follicular development synergistically with gonadotropins. In addition, it stimulates progesterone secretion, through the ovarian upregulation of LH (luteinizing hormone) receptor expression and repression of 20alpha-hydroxysteroid dehydrogenase (20alpha-HSD) . An excess of prolactin, however, blocks follicle development both directly and through the inhibition of GnRH (gonadotropin-releasing hormone) secretion and the LH pituitary response to GnRH stimulation, thus reducing LH pulse amplitude and frequency, finally leading to a PCOM (polycystic ovarian morphology) phenotype . In addition, with excess prolactin, the inhibitory effect of prolactin prevails over aromatase activity, thus leading to low ovarian estrogen production, and a consequent ovarian hyperandrogenism . In addition, prolactin receptors have been documented in the adrenals, and it has also been suggested that hyperprolactinemia could augment adrenal androgen secretion .

Prolactin, therefore, needs to be measured in the diagnostic approach to PCOS, and the discovery of hyperprolactinemia should lead to etiological investigations to exclude all the causes of hyperprolactinemia such as prolactinoma, drugs, hypothyroidism, chronic renal failure, cirrhosis, chest wall lesions, or breast stimulation . In all these cases, the treatment of hyperprolactinemia can lead to an improvement and eventually the remission of PCOS .

However, a macroprolactinemia (excess of “big prolactin” and/or “big-big prolactin”) may be found forming a false hyperprolactinemia, or a true hyperprolactinemia secondary to PCOS. Does the question persist as to how PCOS causes hyperprolactinemia? . The literature on this topic is scarce. However, some data suggest that the slightly high levels of prolactin frequently found in PCOS women are secondary to a decrease in a dopaminergic tone which would also be responsible for the increase in LH levels . Another hypothesis suggests that PCOS could cause hyperprolactinemia because it induces relative hyperestrogenemia . In fact, various experimental studies have shown an increase in the secretion of prolactin under the action of estrogens . Whether the treatment of this “functional form” of hyperprolactinemia in PCOS is recommended to manage the signs and symptoms of PCOS is still a matter of debate.

There are also no clear indications of how to discriminate between mild organic forms of hyperprolactinemia, in particular microprolactinoma, which is the most frequent cause of organic hyperprolactinemia in women of fertile age, from this functional form of hyperprolactinemia in PCOS. A recent study suggested using LH levels, which are reduced in microprolactinoma and normal/high in the functional form . In the same study, the authors also suggested considering prolactin levels in the circulation to differentiate between these two forms. In particular, the authors demonstrated that patients harboring microprolactinoma had significantly higher prolactin levels compared to patients with the functional form (median level of prolactin 95.4 ng/mL vs 49.2 ng/mL, p < 0.0001). They suggested a prolactin threshold of 85.2 ng/mL to distinguish between these two forms with 77% sensitivity and 100% specificity .

To conclude, prolactin should be measured in the circulation in the diagnostic approach to PCOS and, in the case of true hyperprolactinemia, specific treatment should be considered to lead to improvements and eventually the remission of PCOS.

PCOS—Secondary to thyroid disorders

Thyroid hormones participate in the regulation of follicular development and modulate the production of ovarian steroids. Studies in rats have demonstrated that triiodothyronine (T3) amplifies proliferative FSH (follicle-stimulating hormone) action in the granulosa cells (GCs) of the preantral follicle, stimulating its transition to the small antral follicle. In addition, T3 inhibits GC apoptosis . Human studies have also demonstrated that thyroid hormones induce estradiol and progesterone ovarian synthesis, acting directly on GCs . These effects guarantee the maturation of ovarian follicles and the formation of ovulatory oocytes. On the other hand, thyroid hormone disorders in terms of both defect or an excess, alter folliculogenesis and both estradiol and progesterone ovarian synthesis . In addition, hypothyroidism decreases sex hormone binding globulin (SHBG) synthesis and secretion by the liver, with a consequent increased peripheral bioavailability of estrogens and, particularly, androgens, thus producing a condition of “functional” hyperandrogenism . On the other hand, hyperthyroidism causes hyperandrogenism through an increased ovarian production rate of testosterone and androstenedione by a direct stimulatory effect . In addition, an altered GnRH-induced LH secretion is involved with significantly higher LH levels in both the follicular and luteal phases of the menstrual cycle . All these alterations contribute to producing a form of PCOS secondary to thyroid disorders.

Primary hypothyroidism may also cause PCOS through an increase in prolactin which is due to pituitary stimulation by TRH (thyrotropin-releasing hormone) and to peripheral insulin resistance and the subsequent hyperinsulinemia which frequently follows hypothyroidism (see paragraph “PCOS secondary to severe insulin resistance states” for more details) . In addition, increased TSH (thyroid-stimulating hormone) levels in primary hypothyroidism may promote collagen deposition and overall increased ovarian volume, thus contributing to the production of the PCOM phenotype which can revert to normal following thyroxine replacement therapy . The resumption of regular menses and normalization of total and free testosterone may also occur in PCOS after euthyroidism has been achieved .

Many studies also show a higher prevalence of Hashimoto’s thyroiditis in women with PCOS concerning the general population, however, the mechanisms behind this association are not clear, although a genetic susceptibility has been hypothesized . A high estrogen-to-progesterone ratio owing to anovulatory cycles, as well as high androgens are suspected to enhance the autoimmune response in PCOS . A vitamin D deficiency, which frequently occurs in PCOS, may also be involved in the pathogenesis of Hashimoto’s thyroiditis . Finally, obesity, which is associated with most PCOS women, also plays a role through increased leptin levels which aggravate autoimmunity by preferentially inducing effector T-cells and down-regulating regulator T-cells . The few data available also suggest an association between PCOS and hyperthyroidism due to Graves’ disease, although to a lesser extent to the association between PCOS and Hashimoto’s thyroiditis .

Interestingly, PCOS is also frequently associated with increased TSH levels without primary hypothyroidism. The pathophysiological link is probably the insulin resistance that is common in PCOS, particularly in obese PCOS. Through undefined mechanisms, insulin resistance leads to decreased type 2 deiodinase activity at the pituitary level, resulting in the reduced intracellular availability of T3 in thyrotrophic cells resembling a “state of hypothyroidism” and, as a consequence, an increase in TSH levels . Increased leptin levels in obese PCOS may also contribute to increasing TSH by increasing TRH secretion from the hypothalamus via activating the Janus kinase-2/signal transducer and activator of the transcription 3 factor . Increased TSH levels may aggravate obesity in PCOS, by stimulating adipogenesis [19-21], and may promote the adipose production of cytokines, particularly interleukin-6, through the action of TSH receptors on adipocytes . This proinflammatory state may increase insulin resistance and create a vicious circle .

TSH and thyroid hormones in the circulation, therefore, need to be measured in the diagnostic approach to PCOS. In the case of hypo or hyperthyroidism, PCOS should also be reevaluated after the underlying thyroid dysfunction has been appropriately treated.

PCOS—Secondary to nonclassic congenital adrenal hyperplasia due to 21-hydroxylase deficiency

21-NCAH is a common autosomal recessive disease with a prevalence of between 1:1000 and 1:2000 in the general population which, in females, is frequently characterized by a PCOS-like phenotype, due to the increased adrenal androgen secretion driven by increased pituitary ACTH (adrenocorticotropic hormone) synthesis to overcome the block in cortisol biosynthesis . The known prevalence of this disease among the PCOS population is between 1% and 3% , but increases to 33% in PCOS women with basal 17-OH progesterone (17-OHP) in the follicular phase of the menstrual cycle of ≥ 2.00 ng/mL measured by routine immunoassay . Therefore, 21-NCAH should always be considered in the diagnosis of PCOS. However, 21-NCAH does not have specific phenotypic features that can help the clinician to differentiate it from PCOS . Hormone measurements, therefore, represent the only way to guide the diagnosis and to justify CYP21A2 genetic analysis for confirmation. Experts consider a basal immunoassay17-OHP of < 2.00 ng/mL in the follicular phase of the cycle as an effective threshold to rule out 21-NCAH from PCOS since it has a negative predictive value close to 100% . However, the frequency of false-positive cases with a basal 17-OHP of ≥ 2.00 ng/mL is high . Therefore, with a basal 17-OHP in the follicular phase of the cycle of ≥ 2.00 ng/mL a hormonal analysis of secondary levels is needed. The _1-24 ACTH test is the classically recommended secondary test and, recently, the immunoassay cut-off ACTH-stimulated 17-OHP of 6.77 ng/mL was demonstrated to markedly decrease the frequency of false-positives, providing 80.0% specificity and 95.2% sensitivity in differentiating between 21-NCAH and PCOS . Interestingly, using the combined response of 17-OHP ≥ 6.77 ng/mL and cortisol ≤ 240 ng/mL with the ACTH test, a specificity of 100% was obtained . However, in the same study, using liquid chromatography-tandem mass spectrometry (LC-MS/MS) for the fasting blood test instead of ACTH was shown to differentiate between 21-NCAH and PCOS in subjects with an immunoassay of 17-OHP ≥ 2.00 ng/mL in the follicular phase of the cycle. An LC-MS/MS-measured 17-OHP of ≥ 1.79 ng/mL was shown to perform similarly to the immunoassay ACTH-stimulated 17-OHP, with a specificity of 76.9% and a sensitivity of 90.9% . In addition, basal LC-MS/MS measured 21-deoxycortisol ≥ 0.087 ng/mL was found to discriminate 21-NCAH from PCOS with a 100% sensitivity. This thus suggests that this is the best biochemical method for excluding 21-NCAH in a population of PCOS with a positive first hormonal screening and therefore, to completely rule out the chances of an incorrect 21-NCAH diagnosis within a population of PCOS. In addition, the combination of basal LC-MS/MS 17-OHP ≥ 1.79 ng/mL with corticosterone levels of ≤ 8.76 ng/mL provided 100% specificity .

To conclude, 21-NCAH should always be considered in the diagnosis of PCOS due to the high prevalence of this secondary form of PCOS. However, since it is impossible to differentiate between 21-NCAH and PCOS on clinical grounds, laboratory support plays a key role in the differential diagnosis between these two diseases. The simultaneous measurement of 21-deoxycortisol, 17-OHP, and corticosterone by LC-MS/MS is a potentially valid and cost-effective biochemical method for diagnosing 21-NCAH in a PCOS population with a positive first screening test, i.e., 17-OHP immunoassay in the follicular phase of the menstrual cycle of ≥ 2.00 ng/mL. Overall, this tool shows the maximum combination of sensitivity and specificity and should be preferred in clinical practice to the ACTH test, as it prevents the direct and indirect costs associated with test procedures and related patient stress. If LC-MS/MS is unavailable, the combined response of 17-OHP and cortisol to ACTH testing should be considered.

PCOS—Secondary to cushing syndrome

Cushing’s syndrome (CS) is a rare disorder, with an annual incidence of 2–3 cases/million inhabitants and, in females, it is frequently characterized by hirsutism, acne, menstrual irregularity, infertility, hyperandrogenemia, and, therefore, with a PCOS-like phenotype. Although CS is clinically unmistakable when full-blown, the diagnosis can be challenging, particularly in milder cases. Accordingly, a recent study demonstrated that out of 26 women with CS due to pituitary corticotroph tumor (termed Cushing’s Disease-CD), 50% had previously been diagnosed with and treated initially for PCOS . In addition, a recent survey showed that the gynecologist was the first referring physician for 46% of 176 patients with CD . However, another survey demonstrated that only 17% of endocrinologists and 6% of gynecologists biochemically screened PCOS referrals for CS . This thus suggests that PCOS secondary to CS may be misdiagnosed and the specific treatment delayed, with consequent irreversible metabolic, cardiovascular, and bone damage.

Hypercortisolism inhibits GnRH pulsatility and decreases gonadotropin responsiveness to GnRH, thus reducing LH and FSH secretion from the pituitary . In addition, it decreases the ovarian LH-receptor level thereby impairing LH action at the ovary , and directly inhibits ovarian estradiol and progesterone production . Moreover, hypercortisolism increases the level of reactive oxygen species (ROS) beyond the physiological range, resulting in cell cycle arrest and apoptosis in follicular oocytes . Overall, these effects lead to an impairment in follicular development, the formation of atretic antral follicles, anovulation, menstrual irregularities, and in some cases infertility . CS also causes functional hyperandrogenism by reducing the hepatic synthesis of SHBG . In addition, all the forms of CS are characterized by excess adrenal androgen secretion and therefore true hyperandrogenism, characterized by increased circulating levels of dehydroepiandrosterone (DHEA), its sulfate conjugate (DHEA-S), and androstenedione (A4), which have little androgenic activity but act as precursors for the peripheral synthesis of more potent androgens (i.e., A4 in testosterone) . It has been suggested that when hypercortisolism is mild-moderate, gonadotropin stimulation of the ovaries is maintained, whereas when hypercortisolism is severe hypogonadotropic hypogonadism develops. In such cases, hyperinsulinemia as a consequence of peripheral insulin resistance may act as a co-gonadotropin stimulator of the ovaries, thus maintaining ovarian steroid output and contributing to the development of PCOS .

Patients with CS present variable clinical manifestations depending on the degree and duration of hypercortisolism and probably on glucocorticoid receptor sensitivity . The florid phenotype is generally easy to recognize, but in many cases, the picture is much less clear and very similar to a true PCOS, especially in patients with mild and cyclic hypercortisolism . The specific biochemical analysis, therefore, needs to be performed in the diagnostic approach to PCOS to exclude cortisol excess. The biochemical diagnosis of hypercortisolism is based on the demonstration of inappropriate cortisol secretion with the loss of its physiological negative feedback. Measurement of cortisol in more than one 24-h urinary sample and/or the low-dose dexamethasone suppression test (DST), and/or late-night salivary cortisol are recommended as first-line screening tests. The positivity of two of these first-line screening tests indicate suspected hypercortisolism and thus justify the execution of secondary tests, in particular high-dose DST, the corticotropin-releasing hormone test, or Desmopressin (DDAVP) stimulation test .

PCOS—Secondary to acromegaly

Acromegaly is a chronic and rare multisystem disease, with an estimated annual incidence of three cases per million. In females, it is frequently characterized by menstrual irregularities and infertility, principally due to anovulation mainly for pituitary reasons, such as the destruction or compression of gonadotroph cells by the GH (growth hormone) secreting adenoma or hyperprolactinemia derived from tumor co-secretion or the stalk effect (pituitary stalk compression) . However, acromegaly can also be frequently accompanied by a PCOS-like phenotype which is derived from the direct effect of excessive GH/IGF-1 (insulin-like growth factor 1)secretion on the ovaries and/or from insulin resistance secondary to GH excess . In fact, high GH/IGF levels promote an increased synthesis of ovarian androgens through the activation of both type I and type II IGF receptors . In addition, hyperinsulinemia, which compensates for insulin resistance, stimulates androgen production by theca cells (TCs) and reduces SHBG hepatic synthesis, thus increasing the peripheral bioavailability of androgens (see paragraph “PCOS secondary to severe insulin resistance states” for more details).

The florid acromegaly phenotype is generally easy to recognize, but in many cases, the picture is much less clear . Blood measurement of IGF-1, therefore, needs to be performed in the diagnostic approach to PCOS to exclude acromegaly. The discovery of high circulating IGF-1 level according to age reference ranges, justify the execution of secondary tests, i.e., GH measurement during 75-g oral glucose tolerance test (OGTT) .

PCOS—Secondary to severe insulin resistance states

Clinicians who deal with PCOS should always consider the existence of a form of PCOS secondary to a severe state of insulin resistance (SSIR), whose prevalence was recently described to be 1.6% . The reported causes of SSIR are primary defects in insulin signal transduction or adipose tissue dysfunctions due to lipodystrophy or, more frequently, severe obesity . Almost all women with SSIR develop a secondary form of PCOS and present acanthosis nigricans, which are considered the clinical marker of severe IR (insulin resistance), due to a condition of “partial IR” . This condition means that only some tissues are insulin resistant, particularly those involved in the metabolic effects of insulin (skeletal and cardiac muscle and adipose tissue), whereas others such as the pituitary and adrenal glands, the ovaries, skin, and to some extent the liver, maintain a level of insulin sensitivity and, therefore, are exposed to the biological effect of hyperinsulinemia, which compensates IR . Hyperinsulinemia in females produces hyperandrogenism by enhancing the pituitary response of LH to GnRH, by upregulating ovarian LH as well as type 1 IGF and hybrid insulin/type 1 IGF receptors, and by directly increasing ovarian androgen synthesis through the stimulation of cytochrome P450c17α activity . Hyperinsulinemia also increases IGF-1 expression and downregulates IGFBP (IGF binding proteins) production in the ovary, leading to a local increase in free IGF-1 and IGF-2 levels. Hyperinsulinemia may also increase adrenal androgen production through a direct effect on cytochrome P450c17α and/or an increased sensitivity of adrenal to ACTH . Hyperinsulinemia also decreases the hepatic synthesis of both SHBG and IGFBP1, thus increasing the bioavailability of both androgens and IGF-1 and IGF-2 . Finally, hyperinsulinemia causes premature follicular atresia and antral follicular arrest, leading to the development of polycystic ovarian morphology and anovulation . Supported biochemical biomarkers of SSIR are a fasting insulin value above 20.9 μg/mL (150 pmol/L) and or a peak insulin value on OGTT above 209 μg/mL (1500 pmol/L) when diabetes is absent and the BMI is below 30 kg/m ^2. When diabetes is present with an absolute insulin deficiency there is a general consensus that an exogenous insulin requirement of > 3 U/kg of body weight per day is an indicator of SSIR . Unfortunately, there are no clear indications when SSIR is suspected in partial β-cell compensation and/or with a BMI of above 30 kg/m ² .

To conclude, clinicians who deal with PCOS should be alerted to the forms that are secondary to SSIR as they are not as uncommon as might be expected. The diagnosis of these forms remains predominantly clinical, with a focus on clinical hallmarks such as acanthosis nigricans, which are always present, and possibly the lipodystrophy phenotype, and supported by biochemical biomarkers.