Heir Apparent: A Multi-Faceted Look at Congenital Adrenal Hyperplasia

From September 2015, Endocrine News reached out to a basic scientist, a clinical scientist, and an endocrine clinician to discuss the research, treatments, and science behind one of the most common inherited metabolic disorders, congenital adrenal hyperplasia (CAH).

Basic Researcher

Nils Krone, MD, FRCPCH, Centre for Endocrinology, Diabetes and Metabolism, University of Birmingham, Birmingham, UK

Clinical Researcher

Richard J. Auchus, MD, PhD, Professor and Endocrinology Fellowship Program Director
Division of Metabolism, Endocrinology, and Diabetes; Department of Internal Medicine,
University of Michigan, Ann Arbor, Mich.

Clinical Practitioner

Nicole Reisch, MD, Head of CAH outpatient clinic and research group, clinical and translational research team leader at the Medizinische Klinik und Poliklinik IV, Klinikum der Ludwig-Maximilians-Universität, Ziemssenstr, München, Germany

Congenital adrenal hyperplasia (CAH) ranks among the most common inherited metabolic disorders, with the classic form (i.e., 21-hydroxylase deficiency) affecting about one in 15,000 newborns. It is characterized by a complex imbalance of steroid hormones, with reduced cortisol and aldosterone synthesis and excessive adrenal androgen production. Neonatal screening programs in most industrialized countries now enable early diagnosis in classic CAH and timely introduction of hormone substitution treatment. While this reduces life-threatening salt-wasting crises in newborns, hormone substitution over a lifetime poses new challenges in adult patients.

In this Tri-Point article, a basic researcher provides insights from model organism for altered adrenal steroid production, a clinical researcher discusses medical treatment options and challenges for adult patients with CAH and a clinical practitioner highlights fertility related challenges in adult CAH patients.

Basic Researcher’s Perspective — Nils Krone, MD, FRCPCH*

HIGHLIGHTS
· Classic and non-classic presentations of congenital adrenal hyperplasia are caused by a continuum of impairment of steroid enzyme activity.
· In vitro expression analysis of genetic variants has greatly contributed to the understanding of the molecular pathogenesis of all forms of CAH.
· Murine models have improved the understanding of the pathophysiology of CAH; in particular the discovery of impaired metanephrine synthesis in 21-hydroxylase deficiency.

Congenital adrenal hyperplasia (CAH) is a group of autosomal recessive disorders leading to impaired glucocorticoid biosynthesis. The clinical presentation of different CAH forms depends on the underlying enzymatic defect. Deficiencies of 21-hydroxylase (CYP21A2) and 11β-hydroxylase (CYP11B1) only affect adrenal steroid hormone biosynthesis, whereas 17α-hydroxylase (CYP17A1) deficiency and 3β-hydroxysteroid dehydrogenase type 2 (HSD3B2) deficiency also impair gonadal steroid production. P450 oxidoreductase deficiency (PORD) presents with apparent combined CYP17A1-CYP21A2 deficiency. In addition, PORD also causes skeletal malformations and severe genital ambiguity in both sexes. Traditionally, three further enzymatic deficiencies are classified within the CAH group. Steroidogenic acute regulatory protein (StAR) deficiency leads to a lipid adrenal and gonadal lipid accumulation resulting in Lipoid CAH (CLAH). P450 side-chain cleavage (CYP11A1) deficiency resembles the CLAH clinical presentation; however, patients have normal-sized or absent adrenals. Aldosterone synthase (CYP11B2) deficiency manifests with isolated mineralocorticoid deficiency and normal glucocorticoid synthesis.

Genetics of 21-Hydroxylase Deficiency

In clinical practice, the term CAH is often synonymously used for 21-hydroxylase deficiency (21OHD). 21OHD is caused my mutations in the CYP21A2 gene and accounts for about 95% of CAH cases in the general population. The majority of CYP21A2-inactivating mutations are caused by intergenic recombinations within the RCCX module. CYP21A2 gene deletions and CYP21A1P/CYP21A2 chimeric genes account for approximately 20% to 25% of 21OHD alleles in most Caucasian populations. About 70% to 75% of 21OHD disease-causing alleles originate from gene conversions with the CYP21A1P pseudogene, leading in small genetic transfers and single point mutations. The most common CYP21A1P -derived mutations include seven point mutations, a deletion of eight base pairs in exon 3, and the E6 cluster. In addition to these common mutations, more than 150 pseudogene-independent mutations have been reported. The majority of these mutations occur sporadically and have only been reported in a small number of patients.

Genotype-Phenotype Correlation

Soon after the characterization of the CYP21A2 gene and the detection of disease-causing mutations, in vitro expression studies using insect and mammalian cells as well as yeast systems have been employed to assess the residual in vitro 21-hydroxylase function of common CYP21A2 mutations. These residual activities have built the framework for establishing the genotype-phenotype correlation in patients with CAH. This has been widely adopted to report population studies on 21-hydroxylase deficiency. Mutations are categorized into different mutation groups resembling: Null (mutations predicting absent in vitro activity), A (intron 2 splice site mutation), B (mutations such as the p.I172N mutation and mutations with 1% to 10% in vitro residual enzyme activity), C (mutations such as p.P30L, p/V281L, and p.P453S or above 20% to 30% in vitro 21-hydroxylase activity).

Overall, about 65% to 75% of patients with 21OHD are compound heterozygote. Commonly, the clinical phenotype is determined by the mildest CYP21A2 mutation, consequently allowing for the highest residual enzyme activity. Patients with Null and A genotypes have a high likelihood to present with salt-wasting (SW) CAH, whereas two-thirds of patients with group B genotypes present with simple-virilising (SV) CAH. Patients with group C genotypes commonly have non-classic (NC) CAH. Overall, genotype-phenotype data have supported genetic counselling; however, the best correlation has been found with the degree of renal salt loss. A limitation of pure in vitro enzyme analysis is the fact that pathophysiologic changes such as hyperplasia of the adrenal cortex cannot be modelled. This, in addition to other disease modifiers, might explain some of the observed variability of the expressed phenotype in patients with partially inactivating CYP21A2 mutations.

The same principle establishing the in vitro activity of genetic variants has been applied in rarer deficiencies leading to CAH. Over the last five to 10 years the principle of complete and partial mutations has been also described in all other forms of CAH where classic and non-classic presentations are now well established.

Animal Models of CAH

A naturally occurring Cyp21a2-deficient mouse was identified soon after the cloning of the human CYP21A2 gene. Further molecular characterization demonstrated loss of the active Cyp21a1 gene by unequal crossing over between the active Cyp21a1 gene and its pseudogene (Cyp21a2-p) leading to a chimeric Cyp21a1-Cyp21a2-p gene including a partial deletion of Cyp21a1. This chimeric gene carries several pseudogene-derived point mutations, missense mutations, and a nonsense mutation. The in vitro expression analysis of these mutations demonstrated lack of function of this chimeric gene. Overall, the Cyp21a2-deficient mice have been proven to be difficult to maintain postnatally, limiting their use as translational model. Another limitation modelling CAH is the lack of adrenal 17-hydroxylase expression in mice and consequently the lack of androgen excess in the 21-hydroxylase deficient mouse. However, further analysis of adrenals and metanephrine concentrations in these mice demonstrated severe impairment of adrenomedullary function in 21-hydroxylase deficiency. This led to further exploration of the human pathophysiology of CAH showing similar alterations of the adrenal gland in human patients with CAH. Interestingly, the severity of adrenomedullary function is closely correlated with the severity of the CYP21A2 genotype in humans.

Murine models of other CAH forms have been developed. The 11-hydroxylase (cyp11b1) null mouse showed glucocorticoid deficiency and mineralocorticoid excess, adrenal hyperplasia, and mild hypokalemic hypertension as well as glucose intolerance. Remarkably and contrary to humans, affected female mice were found to be infertile. The murine deletion model of the steroidogenic acute regulatory protein (StAR) showed a very high similarity to humans affected with the classic form of CLAH due to StAR mutation. Similar to humans suffering from CLAH, these mice also showed a considerable heterogeneity in the presentation and onset of adrenal insufficiency. The exact mechanisms explaining this variability remains to be elucidated. A similar severe phenotype was observed in the murine deletion model of P450 side chain cleavage model (Cyp11a1). These mice survived to term and then soon die after birth from adrenal insufficiency. In addition, SRY-positive mice show features of sex reversal.

Conclusion and Outlook

In times of more powerful genetic analysis methodologies such as next generation sequencing, modelling of genetic variants, and disease modifiers is becoming increasingly important to understand the clinical consequences. In vitro expression analysis of genetic variants has greatly contributed to the understanding of the molecular genetics of all forms of CAH. Mouse models of CAH have the limitation that the murine adrenal cortex does not produce androgens. However, the discovery of adrenomedullary impairment in CAH has been mainly achieved via studying murine models. Overall, it remains important to reflect that models only answer the specific question asked towards the model system. It is likely that the current revolution in genomic engineering techniques such as transcription activator like effector nucleases (TALEN) and CRISPER/ CAS9 systems will help to study the effects of complete and partial mutations inactivating steroidogenic pathways and enhance our understanding of pathophysiologic consequences by employing novel in vitro and in vivo systems.

*Clinical lead for adrenal disorders and disorders of sex development at Birmingham Children’s Hospital. His lab currently focuses on creating novel whole organism models of inborn errors of steroidogenesis by applying genomic engineering methods.

Clinical Researcher’s Perspective — Richard J. Auchus, MD, PhD

HIGHLIGHTS
· Replacing the cortisol deficiency is straightforward, but better treatments for the androgen excess are needed.
· CAH patients develop characteristic intra-adrenal and ectopic adrenal tumors by unknown mechanisms.
· Obtaining long-term outcomes data for all aspects of CAH will be challenging.

Classic 21-hydroxylase deficiency (21OHD): Not One Simple Defect
The physiology of 21OHD fundamentally reflects a block in cortisol synthesis and diversion of cortisol precursors to androgens under ACTH stimulation. Patients with 21OHD also are prone to develop adrenal adenomas, myelolipomas, and adrenal rest tumors, most commonly in the testes. For women with classic 21OHD, genital virilization, interference with gonadal function, and low rates of fecundity add to the challenges. While some consequences might be predictable, simple models do not explain the variable disease manifestations, and current therapies do not mitigate all complications.

Traditionally, 21OHD treatment has been limited to glucocorticoids, mineralocorticoids, salt, and surgery. Replacing the adrenal insufficiency is not difficult, but normalizing the androgen excess usually requires higher and/or more frequent glucocorticoid doses, particularly to blunt the early-morning rise in ACTH and adrenal steroids. These extra doses create compliance challenges and long-term consequences from life-long over-replacement. Genital reconstruction surgery improves many aspects of appearance and function, but the androgen effects on the brain persist, and chronic urogenital difficulties are common.

Outcomes from large 21OHD cohorts have begun to appear. The NIH report of 183 patients with classic and 61 patients with non-classic 21OHD found no standard treatment regimen, androgens mostly above or below the normal range, short stature, low bone mineral density in 37% of adults, and high rates of obesity and the metabolic syndrome. The British ‘CAHaSE’ study of 203 adults also found mostly abnormal steroid values with a tendency for over-treatment, a high prevalence of cushingoid features, low bone mineral density in 40%, obesity in 40%, markedly compromised fertility, and reduced quality of life. Similar studies from several other groups confirm these findings. Can we do better than these unflattering statistics?

Taking a Step Back and Setting a Clinical Research Agenda

What are the steroids and pathways that lead to virilization in the fetus, bone age advancement in the child, and impaired fertility in the adult?

Ordinarily, the major 19-carbon product of the adrenal is dehydroepiandrosterone sulfate (DHEAS). Androstenedione (AD), testosterone (T) and dihydrotestosterone (DHT) are minor adrenal products; however, T and DHT arise from extra-adrenal metabolism of DHEA. DHEAS is paradoxically low-normal and easily suppressed in classic 21OHD, and recent evidence suggests that DHT in the newborn derives from the 5a-reduced “backdoor” pathway without the intermediacy of AD and T.

How should we monitor therapy at different stages of life, and what are our goals?

Because 17-hydroxyprogesterone (17OHP) accumulates proximal to the block in 21OHD, 17OHP is the gold standard for diagnosis of 21OHD and is widely used to monitor treatment. Yet androgens and estrogens—not 17OHP—cause the problems. Although AD and T are used to monitor therapy, estrogen concentrations below the limits of the best commercial assays will advance bone age. Serum 17OHP is a very sensitive measure of control, but values fluctuate wildly during the day in relation to dosing, and 17OHP correlates poorly with downstream metabolites. The onset of adrenarche and puberty complicate management. In adolescents and adults, sex steroids and 17OHP also derive from the gonads, so one must estimate the contribution from the adrenal. Other steroids such as 21-deoxycortisol are more specific for 21OHD than 17OHP, and this analyte has been used to eliminate false-positive results of newborn screening and to improve diagnostic testing for non-classic 21OHD.

Why do some patients fare so well and others struggle?

With the crisp enzymatic defect and the limited number of pseudogene-derived mutations in most patients, genotype-phenotype correlations in 21OHD are strong but not perfect, even among siblings bearing the same mutations. Response to therapy is highly variable as well. Polymorphism in the genes encoding steroid and drug metabolism enzymes and changes in their expression from environmental exposures might be responsible.

Why do tumors develop?

Some studies have shown absence of ACTH receptors in myelolipomas and no correlation of adrenal rest tumors with biochemical measures of control.

What are the preferred timings and procedures for reconstruction surgeries?

New techniques and trends develop as soon as cohorts from the older procedures reach the stage of “long-term” follow-up.

How can we gather evidence to guide parents and patients for a rare disease with so many confounding variables?

Collaborations and prospective protocols are essential.

Treatments on the Horizon

Sustained-release hydrocortisone preparations are being studied. Greater convenience should improve compliance — but will control be better and complications lower? Continuous subcutaneous hydrocortisone infusion makes sense physiologically and has been shown to improve control with lower total dose in a few patients, but is this approach practical? We showed that abiraterone acetate added to replacement hydrocortisone doses normalized androgens in adult women with classic 21OHD, but the study was short. Will androgen biosynthesis inhibitors benefit children with 21OHD? With innovation and broad collaboration among clinical researchers, we can begin to answer these questions and to improve outcomes.

Clinical Practitioner’s Pespective — Nicole Reisch, MD

HIGLIGHTS
· Fertility is reduced in both males and females with classic congenital adrenal hyperplasia caused by 21-hydroxylase deficiency.
· In males with classic CAH, there are two main reasons for reduced fertility and fecundity: First, testicular adrenal rest tumors–adrenal suppressive therapy with dexamethasone can effectively restore fertility, whereas surgery can neither restore fertility nor HPA axis integrity. Second, adrenal-derived androgens lead to a suppression of gonadotropins causing testicular atrophy and impaired spermatogenesis.
· In females with classic CAH, optimal gluco- and mineralocorticoid substitution therapy has been shown to normalize fecundity.
· In non-classic CAH, an increased rate of miscarriages (25%) can be normalized by low dose hydrocortisone therapy.

Fertility-Related Challenges in Females with Classic CAH

Multiple studies have demonstrated reduced fertility in females with classic CAH because of 21-hydroxylase deficiency. The birth rate in females with CAH is substantially reduced compared to the general population. It is estimated that only 25% of women with CAH ever attempt pregnancy and only 10% of those with the severe salt-wasting form. The reasons for this observation are manifold: vaginal insufficiency with dyspareunia, anovulatory menstrual cycles because of adrenal androgen excess, reduced endometrium growth during follicular phase because of adrenal progesterone production, additional ovarian hyperandrogenism (secondary polycystic ovary) because of adrenal-derived androgens in many patients, multiple psychosexual factors, or masculinization of the central nervous system due to prenatal androgen excess. One Study, investigating 35 adult females with classic CAH, showed that seven were homosexual, eight were married, 13 never have had sexual intercourse, and three patients were without any sexual activity. In this cohort 18 out of 22 patients who actually had sexual intercourse reported painful penetration. The sexual function, as measured by the Female Sexual Function Index, was lower compared to controls and substantially lower in those patients with higher Prader stadium. This finding has been confirmed by a Swedish study showing that sexual life and function is perceived as less satisfactory in genotype “null”.

Most studies on fertility in female patients with CAH, however, do not take into consideration the percentage of patients indeed having attempted to conceive. Thereby, these studies do not represent the actual pregnancy rate. Having a closer look on the actual pregnancy rate in women attempting to conceive, there was no difference in women with CAH and the general population (pregnancy rate in classic CAH 91.3% versus 95%) and no difference in women with the salt-wasting (88.9%) and the simple virilizing forms (92.9%) of classic CAH. In fact, in the majority of women, a spontaneous conception (76.2%) was achieved.

Prednisolone administered once every eight hours has been shown to effectively suppress progesterone in the follicular phase and to improve receptibility of the endometrium. The treatment target for women attempting to conceive is a follicular-phase serum progesterone <2 nmol/l (<0.6 ng/ml). Prednisolone might possibly be superior to hydrocortisone in this respect, however evidence-based data are missing. As dexamethasone passes the placenta, it should be substituted by hydrocortisone or prednisolone to avoid adrenal suppression and possible negative metabolic effects on the fetus. Dexamethasone should only be given if prenatal therapy of the fetus is wanted and warranted. It should only be implemented within the setting of a clinical trial approved by an ethics committee. During pregnancy, renin concentration physiologically increases about five-fold. In classic CAH with and without mineralocorticoid deficiency substantially elevated renin concentrations compared to healthy controls have been described. An increase of fludrocortisone in salt-wasting CAH or the addition of fludrocortisone in simple virilizing CAH may therefore be necessary and reasonable. Hoepffner et al. described a normalization and improved conception after the addition of fludrocortisone.

Fertility-Related Challenges in Males with Classic CAH

For a long time, fertility issues in CAH have focused on females whereas males have received much less attention. In recent years, however, it has become obvious that maintaining fertility in males appears to be even more of a problem than in females. There are no data on sexuality in males with CAH. However, fertility in males is substantially reduced compared to the general population. There are two main reasons for subfertility in males. Firstly suppression of gonadotropins by either adrenal-derived androgens (and their conversion to estrogens) in case of under-replacement of glucocorticoids, or by external glucocorticoids in case of over-replacement. Both mechanisms lead to impaired spermatogenesis and testicular atrophy.

The second major reason is testicular adrenal rest tumors. The pathogenesis of these tumors is not yet completely understood. They are thought to arise from adrenocortical remnants that have descended with the testes during fetal life or from postnatally reprogrammed Leydig cell precursors. TARTs are usually bilateral and always benign. The tumors have been found to express adrenal-specific markers such as MC2-receptor, ATII-receptor, and CYP11B1. They also have been shown to be ACTH-dependent to a certain degree and thus are potentially reversible with the adrenal-suppressive therapy, dexamethasone. They are located in the rete testis and impair spermatogenesis by either mechanical obstruction of the tubuli seminiferi, or they may interfere with the local and/or systemic hormone milieu by their own hormone production. So far, no correlation of TART development and common measures of disease control in CAH has been shown. Still it is very likely that disease control plays a major role in TART growth. As surgical removal of TART has been shown to restore neither fertility nor HPA axis, surgical treatment of TART is only recommended for discomfort and pain-relief and always should be testis-sparing.

Fertility-Related Challenges in Females with Non-Classic CAH

The miscarriage rate in women with non-classic CAH has been shown to be substantially increased (25% vs 6% in the general population). However, it can be normalized with low to moderate doses of hydrocortisone prior to and during pregnancy, therefore it is recommended at these times.