Puberty underlines an important physical and psychosocial period of life where an individual develops secondary sexual characteristics and attains reproductive capacity. Puberty is signaled by the reactivation of the hypothalamic-pituitary-gonadal (HHG) axis, which triggers the release of kisspeptin and the onset of pulsatile hypothalamic gonadotrophin-releasing hormone (GnRH) secretion, which in turn drives pituitary gonadotrophin synthesis and downstream gonadal steroid secretion [
1]. The precise mechanism triggering the reactivation of the HHG axis is not fully understood; however, GnRH neurosecretory activity and thus pubertal timing appear to be partly controlled by complex neuroendocrine pathways gathering genetic, nutritional, hormonal, metabolic, and environmental signals.
Adiposity and pubertal timing
Over the past decades, there has been a worldwide trend towards younger ages of pubertal onset and menarche in girls [
2,
3]. Nowadays, mean age at puberty start is set up below age 10 years, representing an advancement of almost 3 months per decade from 1977 to 2013 [
2]. In contemporary societies, the worldwide rise in childhood overweight/obesity appears to play a key role in the global decrease of the age at puberty. The pivotal role of nutrition and adiposity on pubertal timing has been known since nearly five decades ago, where Frisch and Revelle framed the “critical weight hypothesis” as a key determinant of pubertal start in girls [
4]. Several cross-sectional and longitudinal studies have linked childhood adiposity with earlier pubertal onset, especially in girls [
5,
6]. In some populations, the trend towards early puberty may be even more pronounced, in parallel with rapid gains of body weight [
7], and has a sexual dimorphism, being more noticeable in girls [
8]. This important rise cannot be explained by other environmental factors such as exposures to endocrine disruptors [
9]. Also, a sudden weight gain over a short period of time associated to the lockdown for the coronavirus pandemic has been shown to associate to an increased incidence of precocious and accelerated puberty in Italian girls [
10]. The concept that fat mass in childhood is linked to pubertal timing has recently been endorsed by longitudinal data from >2000 English girls, showing that more fat mass in childhood is followed by an earlier pubertal growth spurt and earlier pubertal completion [
11]. Also, recent genome-wide association studies (GWAS) in humans have identified body mass index (BMI)-increasing alleles that associate with earlier age at menarche, pointing toward genetic co-regulation [
12]. In addition, Mendelian randomization studies support a causal effect of increasing childhood BMI on the risk of early menarche (<12 years) [
13]. Despite these strong epidemiologic and genetic links, the precise mechanism(s) underlying obesity-related early pubertal onset have remained elusive until the discovery of kisspeptin almost two decades ago, connecting the metabolic cues derived from adipose tissue and the regulation of GnRH secretion. It was shown that leptin, which has a permissive role in puberty onset, is able to up-regulate kisspeptin secretion in the hypothalamus, which in turn regulates the pulsatile secretion of GnRH [
14]. Recently, a central ceramide signaling pathway has been unveiled as a novel mediator of obesity-induced early puberty in female rats [
15]. Indeed, reduced signaling by ceramidase and also by AMP-activated protein kinase (AMPK) in the hypothalamus appears to link energy status and puberty-reproduction [
15,
16]. Adiponectin — an adipokine with insulin-sensitizing and cardiovascular-protective properties — signals through its own transmembrane receptors to raise intracellular ceramidase activity, preventing the accumulation of unfavorable ceramide, for example, in the hypothalamus and in the liver [
17].
Early puberty as an adaptive response to ectopic fat accumulation in “mismatch” girls
Earlier/faster maturation in girls has been hypothesized to be the clinical expression of an adaptive mechanism through which girls attempt to escape from ectopic lipid accumulation. This accumulation in turn, results from a mismatch between reduced prenatal weight gain (with reduced subcutaneous adipogenesis, and thus with a reduced capacity for safe lipid storage), and augmented postnatal weight gain (with augmented lipogenesis, and thus, an augmented need for lipid storage) [
18]. Such a mismatch may lead to ectopic lipid accumulation, particularly in the liver and viscera (central obesity), the degree of which may be also influenced by (epi)genetic factors [
19]. The endocrine expression of this mismatch tends to be the early development of insulin resistance, whereas its cardiovascular reflection is often a trend towards higher blood pressure starting in early childhood [
18,
20‐
25]. There are close associations in childhood between the aforementioned mismatch and central fat and also between the mismatch and insulin resistance — as judged by homeostasis model assessment insulin resistance (HOMA-IR)- and between central fat and insulin resistance [
24].
In prepubertal girls, the responses to central obesity include also a decrease in circulating sex hormone-binding globulin (SHBG) and adiponectin, which may be followed by an early and amplified adrenarche, with high levels of its marker, dehydroepiandrosterone-sulfate (DHEAS), and by the appearance of pubic (pubarche) and/or axillary hair, acne and pubertal odor before age 8 years [
26‐
28]. These responses can be viewed as being adaptive since they result in accelerations of body growth and maturation that most likely represent a coordinated feedback mechanism to counteract ectopic adiposity [
18]. If the ectopic lipid accumulation continues, then girls may develop another acceleration of growth and maturation by activating their gonadotrophic axis, conceivably again in a homeostatic attempt to escape from central adiposity. The reduced adiponectin concentrations may favor a reduced intracellular ceramidase activity and thus ceramide accumulation in the hypothalamus and liver triggering pubertal onset [
15]. These novel insights may largely explain the worldwide trends towards younger ages at puberty start and menarche in girls [
2,
3,
5,
11], which are known to associate to higher levels of delinquent and aggressive behavior and to more susceptibility to negative peer influences [
29], and also to future risk of gestational diabetes [
30], type 2 diabetes [
31], and breast and endometrial cancer [
32].
In girls with early puberty, the presence of a mismatch can be easily estimated by calculating the upward change in
Z-score (or centile) between birthweight-for-gestational-age and BMI at onset of puberty. This “mismatch” hypothesis has now been tested in a cohort of girls with isolated variants of central precocious puberty from a single center in Paris [
33], the majority of whom were found to have experienced an upward mismatch between prenatal and postnatal weight gain [
34].
In the first years after menarche, when adult height is almost attained, the compensatory effect of body growth on central fat accumulation is lost. If the energy balance remains chronically positive, the underpinning drive of ectopic adiposity will also remain, and the endocrine-metabolic responses to this drive (insulin resistance, low adiponectin, and SHBG) will persist, and potentially result in a full-blown phenotype of adolescent polycystic ovary syndrome (PCOS) including luteinizing hormone (LH) hypersecretion which in turn, can drive ovarian androgen excess and oligo-anovulation [
16].
Reduction of ectopic fat in “mismatch” girls with accelerated maturation and in adolescents with PCOS: pilot studies
Previous pilot studies performed by our group in rapidly maturing mismatch girls with precocious pubarche and/or early puberty have disclosed that metformin in monotherapy over a period of 3–4 years (up to 850 mg/day), can reduce central adiposity in viscera and liver [
35‐
37], slow down bone maturation [
38], delay pubertal onset [
39], and decelerate the progression of puberty to menarche [
40,
41] and to adolescent PCOS [
42], while augmenting height gain [
40,
41]. In mismatch adolescents with PCOS, a low-dose combination of spironolactone (50mg), pioglitazone (7.5mg), and metformin (850mg) in three separate tablets (SPIOMET) was recently shown to be capable of reversing the entire PCOS phenotype after only 1 year of treatment, including menstrual irregularities, hyperandrogenemia, and insulin resistance, through decreasing hepato-visceral fat excess [
43‐
46]. Here, we propose to conduct a randomized, placebo-controlled, multicenter study using only half of the SPIOMET tablet, i.e., with half-dose spiomet (mini-spiomet: spironolactone 25mg; pioglitazone 3.75mg; metformin 425mg). Administering a triple combination instead of metformin in monotherapy will allow to decrease the dose of each component and to reduce the treatment period to 1 year. The rationale for using three different medications is that each of those medications targets a distinct mismatch-derived dysfunction.
Spironolactone is a steroidal aldosterone antagonist marketed as diuretic but serves as an anti-androgen at higher doses (up to 200 mg/day). Recently, it has been identified as a potent activator of brown adipose tissue (BAT), and thus as a potential driver of energy expenditure, and aims at fat repartitioning [
47,
48]. Spironolactone was first approved in 1960, and it has been used for heart failure, and for other disorders (primary hyperaldosteronism, essential hypertension, edematous conditions). It is licensed for edema in the pediatric population in Europe at doses of approximately 3 mg/kg. No safety concerns related to the use of spironolactone have been raised since its approval [
49]. In Europe and in the USA, spironolactone has been the anti-androgen of choice in the treatment of hirsutism for decades, with an excellent safety profile [
50]. The only minor side effects reported at high dose (100 mg/day or more) are menstrual irregularities, and to a lesser extent, abdominal pain, polyuria, and dryness of the mouth [
50]. There are essentially no safety concerns when dosed at only 25 mg/day (equal or less than 1 mg/kg/day), as will be in this study. Similarly, epidemiologic data show no evidence of an increased risk of any cancer associated with spironolactone use [
51].
Pioglitazone is a thiazolidinedione (TZD) acting as an insulin sensitiser in adipose tissue, liver, and muscle. It raises circulating adiponectin, a driver of intracellular ceramidase [
15,
52], and also insulin sensitivity via preferentially subcutaneous adipogenesis [
53]. Pioglitazone was first approved in 1999, and in 2006, a fixed-dose combination containing pioglitazone and metformin was registered in Europe (Competact®, Glubrava®). At a low dose (7.5 mg/day), pioglitazone acts as an inhibitor of cyclin-dependent kinase 5 (CDK5)-mediated phosphorylation of peroxisome proliferator-activated receptor rather than as a peroxisome proliferator-activated receptor-gamma activator [
54]. The use of pioglitazone has been questioned due to a purported higher risk for bladder cancer in older men with diabetes. A 10-year prospective study performed by the FDA to evaluate this connection concluded that it was non-existing [
55]; accordingly, this association is considered to have been a “red herring” [
56]. In adolescents with PCOS, low-dose pioglitazone (7.5 mg/day) has an excellent safety profile [
43,
44]; pioglitazone appears to be well tolerated by children since no side effects were identified in children (with autism) receiving a 10-fold higher dose [
57]. Pioglitazone is currently under investigation for a first pediatric indication within the SPIOMET context (see below). Spironolactone, when dosed at 50 mg/day in adults, is considered not to cause a clinically relevant drug-drug interaction with pioglitazone, via inhibition of hepatic CYP2C8, which is the main isoenzyme involved in pioglitazone’s metabolism [
58]. In adipocytes, pioglitazone and spironolactone induce the expression of C-X-C motif chemokine ligand-14 (CXCL14), a chemokine that is released by BAT and protects against insulin resistance [
48]. SPIOMET administration normalizes the low levels of CXCL14 in girls with PCOS, suggesting that CXCL14 may be among the mediators of SPIOMET’s benefits.
Metformin has pleiotropic effects but is generally considered to serve as a net “insulin sensitiser” in conditions of ectopic adiposity with insulin resistance; in addition, it raises AMPK activity, and the circulating concentrations of Growth-and-Differentiation Factor 15 (GDF15), a peptide hormone that reduces hepatic steatosis and raises intestinal glucose utilization thereby promoting weight loss [
59‐
61]. Metformin was first approved in 1959, and since then several FDCs have been approved for the treatment of type 2 diabetes as first- and/or second-line therapies [
62]. Worldwide, metformin is the drug most widely prescribed for the treatment of type 2 diabetes in adults and in children older than 10 years; its use has significantly increased in younger children and adolescents without diabetes, including for early maturation and PCOS in girls [
37‐
41,
63]. Extensive experience has been gathered over the last 60 years related to the clinical use and safety of metformin [
64]. In 2001, the European Medicines Agency (EMA) issued a favorable benefit-risk ratio for metformin that outlines its safety in humans [
62]. The main side effects are gastrointestinal symptoms (~10%), that usually resolve after therapy start [
64]. Lactic acidosis has been only described in cases of renal, cardiac and hepatic failure, or after intentional overdose [
65]. A decrease of vitamin B12 serum levels may occur after long-term treatment but appears to be of no clinical relevance [
66]. The combination of spironolactone and metformin is not associated with a higher incidence of adverse events compared to low-dose spironolactone or metformin in monotherapy [
50]. Based on the EMA Summary of Product Characteristics (
https://www.medicines.org.uk/emc/product/594/smpc#POSOLOGY), the dose range studied for metformin in clinical trials is mainly 200–850 mg/day, with a maximum of 2000 mg/day. Hence, the proposed dose of metformin (425 mg/day) will be in the lower recommended range, assuming that the weight of pubertal girls aged 8–9 years will be >25 kg [
67].