American J o u r n a l of Medical Genetics Supplement 7:20-30 (1990)

Pathogenesis of Mental Deficiency in Trisomy 21 JerBme Lejeune Centre de Recherches Claude Bernard, Gdndtique Hurnaine et Maladies de llntelligence, Znstitut de Progenese, Paris, France ~~

In trisomy 21, pathogenesis of mental retardation is still poorly understood although the knowledge of the genic content of chromosome 21 is steadily increasing. Short of discovering how to silence selectively one of the 3 chromosomes 21, no rational medication can be envisaged before pathogenesis has been unraveled, at least partially. A biochemical scheme of impairment of mental efficiency is presented. Secondarily, the possible deleterious effects of a given gene overdose are discussed. Cu/Zn SOD, cystathionine beta synthase, S loop protein, phosphofructokinase, purine synthesis and adenosine pharmacology, thyroid disturbance, and elevated TSH with low rT3 as well as biopterine metabolism interferences are reviewed. It is observed that the metabolic paths controlled by these genes, although unrelated at first glance, are in fact tightly related by their effects, just as if synteny was in some way related to biochemical cooperation or mutually controlled regulation. Experiments in vitro have demonstrated a peculiar sensitivity of trisomic 21 lymphocytes to methotrexate. From this starting point, systematic research of special sensitivities has begun. Clinical observations and relevant statistical methods allow study of the speed of mental development under various medications. The interest of regulating thyroid metabolism, when needed, is exemplified. Reequilibration of monocarbon metabolism is discussed and the seemingly favourable effect of folinic acid medication in pseudo-Alzheimer complication is presented.

KEY WORDS: mental retardation, methotrexate, monocarbons, thyroid, folic acid, Alzheimer ~

Received for publication August 8,1989; revision received March 5, 1990. Address reprint requests to F’rofesseur JerBme Lejeune, Institut de Progenese 45 rue des Saints Peres, 75270 Pariscedex 06, France.

0 1990 Wiley-Liss, Inc.

INTRODUCTION With upward-slanting eyelids, a little nose in a round face, and incompletely chiseled features, Down syndrome patients look more like children than the usual child does. Every child has short hands with short fingers, but theirs are shorter. All their anatomy is rounded, with no harsh features or stiffness. Their ligaments and muscles have a suppleness producing a tender languor in their posture. This general softness extends even to their character: cheerful and affectionate, they have a special charm easier to cherish than to describe. That is not to say that Down syndrome is a desirable condition. It is a n implacable disorder depriving the children of the most precious quality afforded by our genetic patrimony, the full power of rational thinking. This combination of a tragic chromosomal error with a really attractive nature reveals, in a glimpse, what medicine is all about: to fight against disease and to love the disabled. While pondering over this evening’s talk, I suddenly realized that with all the progress accumulated during the last 30 years, the destiny of the trisomy 21-affected persons has not yet been substantially ameliorated. Remarkable achievements in cardiac surgery and in management of infectious or malignant diseases have greatly extended their life expectancy. But a t the same time, early detection and selective abortion have drastically reduced their rate of survival. Looking a t some statistics, i t seems that for a fewmonth-old trisomic 21 baby in utero the rate of survival up to 10 years of age was possibly greater 30 years ago than it is today. Such a n estimate includes the postnatal dangers: deliberate neglect, denial of life-saving interventions or of simple nutrition, and even direct infanticide; health by death is a desperate mockery of medicine. Let us look a t another terrible and incurable ailment: Alzheimer disease. An enormous effort, worldwide, is very aptly made in its study. The lives of millions depend on the success of this effort. But the gene of the familial form [115]is on chromosome 21; the gene of the precursor of the amyloid substance [11,131] is also on 21 and trisomy 21-affected persons are especially prone to presenile dementia [59,36,52], although there is no excess of Alzheimer disease in their families [9]. Certainly a microscope

Pathogenesis of Trisomy 21 must not be construed into a 1:rystal ball, but I would venture to say that a victory over the neural disturbances resulting from the genic overdose of trisomy 21 would very likely also lead to a cure or to a prevention of Alzheimer dementia. The reciprocal prediction looks much less likely. Could it be t h a t in the absence of a cure for the latter, it looks futile to try to cope ivith the inborn form of mental deficiency? Is a genic averdose never amenable to treatment? So grave a matte].deserves careful discussion.

THE SYMPHONY OF [NTELLIGENCE The message of life can be compared to a symphony: each musician (the genes) reads a score and follows the tempo of the conductor. During a solo, a too-quick riusician (in case of trisomy) could transform a n “ztndante” into a “prestissimo”: the ears will be too s nall and the fingers too short. Conversely, a slow music a n (in the case of monosomy) could change an “allegreti 0’’into a “largo”: the ear will be chiseled and the fingcrs too slender. In both cases, because the musician played a solo, he modified a trait but did not spoil the whole symphony. Hence the type-countertype opposition between trisomy and monosomy [69]. On the contrary, when the ful I orchestra is playing, all the musicians playing in a “tutti,” it does not matter whether the faulty musician a xelerates or slow down; the result will be cacophonic, eiren if he reads his music correctly! Hence the mental deficiency in trisomy as well as in monosomic states: human intelligence is the top of our genetic endowment. Detecting the discordant output musicians is not a n easy task especially when a u hole chromosome is involved as in Down syndrome. Surely, most of the genes do not produce harm when in triplicate, because trisomic children would not survike a t all. Few of the accelerated reactions are dangerous; but how will we detect the culprits among so many innocents! This detective story could be avoided if we knew how to silence a specific chromosomc?without disturbing the others. Let us suppose that a competent car-repair man has received from the factory a 4-cylinder engine equipped by mistake with 5 sp 3rkplugs. He would certainly notice that the engine does not run smoothly. An ignorant man would discard t k is motor, but a n expert would cleverly disconnect the ei:tra plug and thus bring the rhythm to normal. Nature if. that shrewd; she knows how to silence one of the X chroinosomes in female cells, so that the woman with her two X chromosomes is not so much superior to the man who k as only one X and a tiny Y! We are still ignorant of liow this turning-off is achieved. Pending such a “tour de force” applied to chromosome 21 we have to analyse its geni: contents and consider how it could affect neural efficiency. A BIOCHEMICAI, SCHEME As already discussed [72] the functioning of the brain necessitates:

21

An enormous number of components: Some 11 billion neurons. A logical wiring of a considerable length: Some 5,000 km if counted in dendrites and axons and from here to the moon and hopefully back if measuring the neurotubule network inside the neurons. A specific response of the gating system, through chemical mediators acting on appropriate post-synaptic membranes. To meet these 3 requirements, the brain has to synthesize a considerable amount of: Monocarbons for synthesis of chemical mediators and for their subsequent inactivation, and all the methylation pathways. Purine and pyrimidine for RNA and DNA maintenance. Tubuline for the wiring and biopterine for the aromatic hydroxylations of the mediators. It has already been remarked that a block of one step of any of these pathways does produce mental retardation [721. The painful task of unraveling one by one the genes of chromosomes 21 is achieved by 2 methods: The molecular biologist split the DNA and letter by letter deciphers the message encoded in each piece. The biochemists carefully analyse chemical reactions in order to pick out those running too fast in trisomy 21. The results of these 2 convergent approachs can be summarized in a general chemical scheme (Fig. 1) The first column deals with purine synthesis. The second deals with pyrimidine (top) and purine (below) interconversions. The third deals with folate and monocarbon metabolism (top), biopterin and hydroxylases (middle), and methylation (bottom). At the far right end products appear, required for a n informative network (tubilin), a gating process (chemical mediators) and a n insulating system (myelin). These 3 categories are absolute prerequisites for the function of the brain (see [72] for general discussion).

GENES ON CHROMOSOME 21 Superoxide Dismutase The first enzyme assigned to a gene on chromosome 21, superoxide dismutase activity is increased by a factor of 1.5 [l20]. Too much 0,- is transformed into hydrogen peroxide H,O,. Glutathione peroxidase, which turns H,O, into H20,is also increased [1211, although its gene is on chromosome 3. Remarkably a glutathione peroxidase-like gene is located on chromosome 21 [85]. The superoxide ion is required by indolamine oxidases (tryptophane and hydroxytryptophane); biopterines are also involved in these reactions [go]. Experimentally, SOD1 protects the activity of the 5’deiodinase, normally inactivated by the superoxide

22

Lejeune

c

T

P

DOVA

-dcDP

5ALA

A D E N I N E

U R A T E

CYSTEINE

Fig. 1. Biochemical aspects of metabolic basis of intelligence; see text.

ion [%I. Thus, excess of SODI could increase the transformation rate of rT, into inactive T2,thus, contributing to the low rT, level found in trisomy 21 1761. In transgenic mice [51 excess of the CdZn SOD gene produces abnormal neuromuscular junctions reminiscent of those seen in trisomy 21. Note that the production of the superoxide ion by human neutrophils is inhibited by adenosine acting upon a membrane receptor [27]. Thus, excess activity of SODI could be related to: Diminished input of oxydized monocarbons (indoleamine oxydases) and impaired biopterine metabolism. Decrease of rT, (5’deiodinase). Correlative change in neuromuscular junction. Cystathionine Beta Synthase Inactivity of the cystathionine beta synthase (CBS) leads to accumulation of homocysteine not transformed in cystathionine. Homocystinuric children are tall and slender with long, tapering fingers with extra flexion creases, contrasting to the small stature and short fingers lacking some creases of trisomic 21 children. This type and counter-type effect led to prediction of a n anomaly of CBS [701 confirmed 10 years later by the localization of the gene on chromosome 21 [1231 and the dosage effect demonstration [ 191.

In homocystinuria, excess S-adenosyl-homocysteine (SAH) competes with S-adenosyl-methionine (SAM) and inhibits the transmethylases. In trisomy 21, insufficient homocysteine [20j could impair the remethylation pathway via 5 methyl-THF and B12 and slow down the recovery of SAM. A reduced rate of methylation of nicotinamide was demonstrated 30 years ago 1451. Homocysteic acid promotes growth of rats [24], but it remains a n open question how homocysteine availability, which affects thymidine synthetase, could also modify thyroid regulation or growth hormone production. SlOOp. Recently localized on chromosome 21 [2], the SlOOp protein subunit plays a n important role in neural function. Appearing late in the forebrain [23,142], its level is high in the hippocampal region; it increases during the learning process [56] and resembles the neurite extension factor [66]. Modulated by calcium [95], it has also great affinity for phenothiazine [83] and, remarkably, for zinc [71. Zinc has neurotrophic properties [1281. It modulates thyrotropin excretion [601 and modifies the sensitivity of the N-methyl-D-aspartate receptor in the hippocampus [1361. Zinc is reported to be beneficial in Down syndrome: i t enhances neutrophil chemotaxis and immune function [lo]; i t reactivates the serum thymic factor [391 which is low in Down syndrome as well as in hypothyroidism 1371. The increase of Cu Zn SOD and of

Pathogenesis of Trisomy 21

Sloop could very well increlse the requirements of Z n + + .In addition 5‘ nucleotidase, the main producer of adenosine, is also a Zn+‘-reqiiiring enzyme. Beside these interactions, an excess of SlOOp could be deleterious because of its abil t y to disassemble brain microtubules [301. Phosphofructokinase (PJ‘K). Excess activity of PFK could increase fructose-1 6-diphosphate, which is known to accelerate biotin ace1yl-CoA carboxylase, the first step of lipid synthesis. Whether this could be related to obesity is not known. Curiously, Sloop has a q e c i a l affinity for fructose-1,6-diphosphate aldolase 11411, the step following PFK. Could this peculiarity bc another indication of a subjacent biochemical logic to gene localization? The same notion could appljr to the enhancement of PFK activity by NH, producec by AMP deaminase in the “purine cycle” [801, which it8especially important in the brain and probably abnorrial in trisomy 21. Genes Pertaining to P u r i n e Synthesis Slight overproduction and ovc rexcretion of uric acid in Down syndrome patients we1 e recognized long ago [42,93,4].Among the various sttbps ofpurine synthesis, 3 are known to be controlled by g mes on chromosome 21: the third, leading to synthesis c f 5-P-rybosylamine; the fourth, formyl-transferase, eading to N-formylglycineamidine riboside (FGAhl); and the sixth, aminoimidazole synthetase, leading: to 5-aminoimidazole riboside (AIR). Overproduction of purine nec 3ssitates more phosphoribosyl pyrophosphate (PRPP) ,md the first step in the production of its precursor rib1 lose-5-P, by the hexose monophosphate shunt, is accelcrated [121,88,1051. Although the demand on PRI’P and monocarbon metabolism remains moderate, there are reasons to believe that purine metabolism imbalance is worth further investigation. In trisomy 21 erythocytes 1111 and lymphocytes [110], there is a n excess of adenosine deaminase (ADA) and of purine nucleoside phospl- orylase (PNP), in accordance with increased urate excretion. An excess of AMP [63] and ADP [129,81] exists ir erythrocytes, however, with a normal level of ATP. This draws attention to adenosine metabolism, especially because a block of adenylosuccinate lyase producing AMP from adenylosuccinate as well as 5-amino-4imidazole-carboxamide riboside (AICAR) from 4-N-succinocarboxamide-5-aminoimidazoleriboside (SCAIR), which is located on chromosome 22 1651, induces a very severe syndrome [57,58].The ps:Ichotic behavior is quite the counter-type of the happy cl taracter of “easy-going” Down syndrome children. But, in rare cases, severely affected trisomic children exhibit autoagressive behavioL r, biting their fingers, banging their heads, quite reminiscent of children with the Lesch-Nyhan syndrome. In 1 his devastating disease lack of hypoxanthine guanine phosphoribosyl transferase (HPRT) prevents the sal\,age of guanine, hypoxanthine, and xanthine and ne zessitates a n excessive synthesis of purine to cope with this permanent leakage. Some Lesch-Nyhan patients ell Crete AICAR, possibly

23

because of insufficient availability of 10-formyl tetrahydrofolate, too severely required by purine production [1071. Remarkably, AICAR is the step just following SCAIR, the product insufficiently metabolized in adenylosuccinate lyase deficiency. The block of adenylosuccinate synthase in mouse cells produces an excess of GTP [1331, demonstrating the intricacy of these regulations. Adenosine modulates the release of chemical mediator [106,341 a t central and peripheral synapses. Its action (essentially presynaptic) has numerous pharmacological consequences C38,41,28,82,1381. From the experimental data one can foresee many effects which could result from excessive adenosine formation and compare them to frequent trisomy 21 symptoms: A mild deficit of immune reaction, suggested by the dramatic effect of ADA deficiency [681 vs. the classical sensitivity to various infections. A deficiency of the neuromuscular transmission [118,122] vs. severe hypotonia. A deficit of growth hormone secretion [31] vs. short stature with normal sensitivity to growth hormone [31. Inhibition of lipolysis after adrenergic stimulation of the adipocytes 1117,1351 especially in hypothyroidism [92], vs. frequent obesity and thyroid deficiency. Instability of blood sugar levels [17,78,1391 vs. the frequent prediabetic state. Abnormal pupillary reaction [48] vs. the hypersensitivity of the iris to atropine [711. Relaxation of the vascular muscle 1281vs. the livedo reticularis, observed also in homocytinuria, and relaxation of intestinal muscle [28] vs. frequent constipation. The hypersensitivity of fibroblasts to P-adrenergic stimulation [871may also be related to adenosine effect. The importance of adenosine modulation in the cerebellum, in the hippocampus [33], and in the innervation of trigeminal mescencephalic primary afferent neurons from hypothalamus [891 could be compared to the waddling gait, the frequent grinding of teeth, the nystagmus, and, possibly, the slight instability of the thermoregulation in the newborn infant. Similarly, adenosine deficiency could produce instability, irritability, anxiety, or even convulsions, as suggested by the effects of adenosine antagonists like theophylline [106,82,1121 and caffeine [32,29,26,671 or even psychotic behaviour as suggested by the antipsychotic-like properties of adenosine receptor agonists 1511.

B Amyloid Precursor Protein (A41 Different from the locus of the familial predisposition to Alzheimer disease [ 115,151, the p amyloid precursor gene [46] is closer to the centromere (21q21.1) than the “Down syndrome region” (21q23.11, [108,130,131I. Accumulation of amyloid substance in senile plaques is one of the signs of Alzheimer disease [861. It occurs also in Down syndrome and other diseases [16,471, in angiopathy [25], in thyroxin transport defect (transthyretine) [54,84] and in pugilistic dementia [53,1141. The first 28 amino acids of A4, as a free polypeptide,

24

Lejeune

increase the survival of pyramidal neurons ofthe hippocampus cultured in vitro [137]. Accumulation of A4 could result from a n inappropriate utilisation and be a symptom more than a cause of the cerebral impairment.

Thyroid Dysfunction In 1906 de Bourneville was the first to treat hypothyroidism in a trisomic patient by opotherapy [14]. The 2 diseases have always been closely linked and Benda described “thyroid exhaustion” after a series of autopsies 181. Since 1981 [116] a n excess of thyroid stimulating hormone (TSH) has been widely observed (for a general review, see [76]). Aside from the increased frequency of true hypothyroidism, a moderate excess of TSH is found in nearly half of apparently euthyroid patients. Tetraiodothyronine (T,) and 3, 5, 3’-triiodothyronine (T,) remain within normal values. Moreover, a characteristic deficit of 3,3’,5’-triiodothyronine, or reverse T, (rT,) is observed [761. The ratio rT,/TSH (an index of the yield of rT, per unit of TSH) is highly significantly diminished. This rT,/ TSH ratio, highly correlated with T, level normally, is not correlated in trisomy 21 [76]. These new facts demonstrate thyroid dysfunction, especially for rT,, which regulates T4 and T3 1221. Low rTg, possibly related to superoxide dismutase excess (see above), could impair growth hormone stimulation [94] possibly jeopardized also by a n adenosine effect (see below). The intricacy of these regulations is illustrated by the fact that the level of 5’nucleosidase, one of the producers of adenosine, is controlled by thyroxine 11241. Remarkably, thyroid hormone induces gene expression through a responsive element common to retinoic acid 11341; disorder of carotene and vitamin A is frequent in Down syndrome 11271. Monocarbon metabolism, which could play a key role in mental deficiency [721, is strongly connected to thyroid function. Thyroxine increases the input of oxidized monocarbons (-CHO) by activating the 10-formyl-synthase and the output of reduced monocarbons (-CH,) by activating 5-10-methylene-THF-reductase [1191. Simultaneously, thyroxine preserves the monocarbon pool by inhibiting the 10-formyl-THF-dehydrogenase (which eliminates -CHO into CO,) 1623 and by inhibiting the cystathionase [21] (which disposes of homocysteine into cysteine and homoserine). This interference with the methyl carrier is much more interesting, since methionine has antinomic properties [1191. It increases activity of 10-formyl-THF-dehydrogenase (depleting the monocarbon pool) and blocks the 10-formyl-THF-synthase and the 5-10-methylene-THF-reductase (diminishing the input and the output). Hence, thyroxine and methionine metabolisms, both abnormal in trisomy 21, play antagonistic roles in the monocarbon metabolism, and are abnormal also in this disease (see purine metabolism). Tubulin organisation, using GTP [1401 is largely controlled by thyroid function 1911. As already discussed [74], the spindle of the mitotic apparatus and the neuro-

tubules are made of the same building blocks: tubuline. Hence, a cell has to choose either to keep assembling and disassembling tubuline for mitotic machinery or to mount tubuline into a n inner informative circuitry. And this ordinated network is disrupted by neurofibrillary tangles in 3 diseases: Down syndrome, Alzheimer, and hypothyroidism.

Biopterin Metabolism Controlling the production of adrenergic and serotoninergic mediators via hydroxylases, biopterin seems imperfectly regulated in trisomy 21 [491; a low level of BH, in the brain suggests a deficit of dihydrobiopterin reductase (QDPR). A recent investigation [18] showed a n elevated urinary ratio of neopterinhiopterin, a shift already observed in Alzheimer disease 1491. A slight overproduction of GTP together with a lower quinonoiddihydropterine-reductase (QDPR) efficiency could be present. Enzymes controlling folate (a vitamin) and biopterin (a “home-made” cofactor) can partially replace each other, suggesting a kind of fail-free system [741. Both can be eliminated by oxidation into isoxanthopterin. A trisomy 21 child, overeliminating isoxanthopterin [18], was secondarily found by anamnesis to have been at that time suffering a typical regression diagnosed 2 months later a s severe hypothyroidism. Whether thyroid hormones could prevent folate and biopterin elimination via isoxanthopterine is a n open question. EXPERIMENTS IN VITRO In 1985, clinical serendipity paved the way to experimental investigation. PEETERS et al. [96, 971 discovered that leukemic children could not tolerate normal doses of methotrexate if they also had Down syndrome. Toxicity of this inhibitor of dehydrofolate reductase appeared at half the normal dose [99]. This phenomenon has since been amply confirmed [40,98,431. In 1986, we demonstrated [731 that trisomy 21 lymphocytes are twice as sensitive as normal ones to methotrexate. Sensitivity to chromosome breaks is also increased [13,1041. These facts are in accordance with the symptoms of folate deficiency in the leucocytes [441. Generally speaking, it must be remembered that methotrexate is dangerous for the human brain as observed after treatment for leukemia, with or without cranial irradiation [61,109,35,1I. Although time consuming, the mitotic index method allows a first approach to the specific sensitivity of trisomy 21 lymphocytes a s exemplified by the results of PEETERS et al. [ 1011.Methotrexate hypersensitivity is fully confirmed [103,99] and systematic investigation showed that thymidylate synthase and thymidine kinase pathways are normal [1031. Contrasting with this methotrexate toxicity, 6-mercaptopurine (inhibitor of IMP dehydrogenase and also of adenylosuccinate synthase) has, a t low doses, a beneficial effect on the mitotic index [ 1021; aracytine has also a beneficial effect [loll. All these facts point toward a dysregulation of folate metabolism with a dysequ-

Pathogenesis of Trisomy 21 ilibrium between adenosine and guanosine derivatives as well as between purine an11 pyrimidine pathways. IMP dehydrogenase can also be modified by a specific inhibitor: mycophenolic acid ( N PA) [791,a natural modulator: 2-3 diphosphoglycera1.e (2-3 DPG) [771 and reverse triiodotyrosine (r’Il3). In trisomy 21 erythrocytes, 2-3 DPG is low [63,64,81], its synthesis is under thyroxin command [125.1261, and rT3 is also low in trisomy 21 [76]. Pre1iminar:r data show a very tight correlation between the effects 3f these 3 products when tested on the lymphocytes of tne same patient. A possible explanation of the ilisequilibrium observed in purine metabolism could bt! that a small excess of GTP (see biopterin) would accelerate adenylosuccinate synthase and a t the same time liminish the IMP deaminase, making more AMP avai .able. At the same time, disposal of homocysteine by C I S would liberate S-adenosyl-homocysteine hydrolase. These 2 effects could increase adenosine production (see purine synthesis).

THEORETICAL CONSII~ERATIONSAND CLINICAL OBSERVATIONS From these considerations ail heuristic investigation could be directed toward: Controlling the dysthyroidism. Compensating abnormal puiine derivatives. Equilibrating the homocysteinelmethionine pathway. Increasing folate or biopteri I availability.

Clinical Data No systematic attempt has keen made because each child has been treated with thit medication considered the best for his or her persona state. Nevertheless, some data on thyroxine, methionine, and folic acid, however scanty o possibly biased, can be extracted from the observations accumulated with the collaboration of associate Professor M.O. Rethore and Doctors M.C de Blois, C. Pangalos, M. Peeters, M. Prieur, P.M. Sinet, and 0. Raoul a t tke Hospital des Enfants Malades. Thanks to the extraordinarj cooperation of the patients and of their sibs, and Tvith the fully informed consent of their dedicated parents, laboratory andlor clinical examinations are gatkered a t a rate of some 2,000 per year, with a total of sclme 5,000 recorded files. Every 6 months, a patient ie submitted to a psychometric test. His or her personitl progress is compared step by step to the general charl, established long ago on 100 Down syndrome children (Fig. 2) After each period of 6 month j on a given medication, an eventual “inflexion” of the Dersonal curve is calculated according to the expected evolution (see legend of Fig. 2). This parameter expresses the eventual betterment or deterioration of the ptmonal curve of a given child. “Inflexion” has by defini ;ion a mathematical expectation of zero and a n experimental standard deviation of about 0.5. All the assays have been foliowed in this manner. If the mean “inflexion” is statisticmally different from zero,

-

25

this is taken as a n indication that the “treatment” has modified the development which would have probably taken place if no “treatment” had been given. Thyroxin therapy. (Table I) Thyroxin (T,) was prescribed only if TSH was elevated with low T3 or T,. Varying from 3 pglkglday a t 6 months to 1 pglkglday after age 12 years, the dose was adjusted according the ensuing shift in TSH, T3, and T, values. For children less than age 5 years 95 “inflexions” could be calculated. The mean I.&. of this sample being 74.3 2 15.9 the patients could be classified as “gifted” above 70 and “less gifted” below 70. The 36 “less gifted” had a mean inflexion of + 0.454 c 0.709, this value being significantly higher than the + 0.140 t 0.754 observed in 82 equally “less gifted” of the same age receiving no T, (and no methionine). The 59 “gifted” ones had a mean inflexion of - 0.201 c 0.724 which did not differ from the - 0.062 2 0.720 observed in equally “gifted” children not receiving T, nor methionine. The highly significant difference between “gifted’ doing poorly with thyroxine and “less gifted” getting a great benefit of it (t= 4.3; P 0.025) and being insignificantly unfavourable for the “gifted” ones. For children over age 5 years, the same tendency is observed, although not significantly; for 136 “less gifted” (below the mean I.&. of 5 5 ) inflexion = + 0.062 2 0.509 and for the 150 “gifted” (I.&. above 5 5 ) infl = -0.038?0.471. Folk acid medication. (Table 111) One hundred forty-three “inflexions” are available for less than 5-year-old children who received neither methionine nor T,. Sixty-nine received moderate folic acid doses ranging from 5 mg to 35 mg per week and 74 received no folate a t all. The mean “inflexion” was + 0.1514c 0.828 for the folate group against - 0.037 0.649 for the non-folate, the difference being not significant. For children over age 5 years, the same tendency is observed and all the data could be pooled (children from 1year to age 12 years). The 109 folate receivers showed a mean inflexion of + 0.129 2 0.689 and the 172 non-folate a mean of - 0.054 c 0.508. The difference is significant t = 2.56; 0.05 > P > 0.01). A multivariate analysis on a greater sample (now in progress) will be required in order to confirm this seemingly beneficial effect of moderate doses of folic acid and to investigate a n eventual doseleffect relation. Folinic acid medication. [see ref. 751 A trial of medication with folinic acid (5-formyl-tetrahydrofolate) was performed on 39 severely affected trisomic 21 patients [751. Thirty of them had a n infantile psychosis

*

26

Lejeune

chronological I

2

3

4

5

6

7

age ( y e a r s )

8

I

I

9

10

I

11

Fig. 2. A cohort of 100 Down syndrome children was followed from one year up to age 14 years. A psychometric test was administered twice a year (Brunet-Lezine first, Bore1 Maisonny later, and finally Binet-Simon). Mean mental age is reported on the ordinate with chronological age on the abscissa. The local value of standard deviation (S.D.) is given in months. Analysis of more than 700 successives tests (independent of those reported here) demonstrated that each child follows his “personal trajectory,” roughly parallel to the curve of the general mean. For each child fluctuations around his own trajectory are contained inside a corridor with width equal to 1 S.D. of the general population (data not shown). For example a t chronological age 4 the child had a mental age of 3y. His performance 6 months later, Ye, was thus expected to lie on a line parallel to the curves of the mean and of the S.D.;thus, Ye = 3y 3m at chronological age 4y.6m. The observed performance, Yo, being 3y 10m, the “inflexion”of his curve can be expressed as 3~ 10m - 3~ 3m = +1=1.08 6.5m 6.5 This “inflexion” Yo - Ye local S.D. has a mathematical expectation equal to zero. The standard deviation of this parameter was found to be of the order of 0.5, ranging from 0.4 for children between 5 and 12 years of age to 0.7 for children less than 5 years old. For a cohort of children receiving a given treatment, the “mean inflexion” ? S.D. can be calculated (see Tables 11-111). If this “mean inflexion” is more than twice its S.D., this is taken as an indication that the medication has modified the speed of development of the mental age.

and the other 9 suffered from Alzheimer-type regression. Of the 69 assays, 37 were favourable and 32 were not, with no untoward effects. Considering the quasi-inexorable course of these 2 complications, this modest result is nevertheless very suggestive of a really beneficial effect. Unexpectedly, a doseleffect relationship was noted, suggesting the efficient dose being rather high a t some 0.8 to 1 mgikilolday.

CONCLUSION No simple management, short of chromosome turning-off, can be predicted for trisomy 21. The chemical basis of the mental deficiency must be a disruption of a fantastically co-ordinated system. The metaphor of a n orchestra in “concert” was not purely rhetoric. Trisomy 21 is dis-concerting. For each chem-

ically defined disease known to produce mental retardation, one finds a more or less evident anomaly of the sensitive step in Down syndrome. This evidence is, in a sense, encouraging. For example, it could very well be that the regressive infantile psychosis sometimes observed is caused by hypothyroidism that went unnoticed and reequilibrated itself later. The same may be true for purine, pyrimidine, folic acid, biopterine, or methionine metabolism. Surely, we are not yet able to restore the destiny, but we are possibly just in the situation of being able to prevent its worsening. It must be very precisely stressed that this general model is for the moment strictly speculative. Even if the reasoning is sound, it remains to be seen whether the correction of such troubles will, in the long run, alleviate the mental deficiency.

Pathogenesis of Trisomy 21

27

TABLE I. Mean “Inflexion” t S.D. for Untreated and Thyroxine-Treated Trisomy 21 Children (6-Month Period) Age

Trisomy 21

Less than 5 years

No. of 6-month periods untreated 82 thyroxine 36 untreated 61 thyroxine 59 untreated 62 thyroxine 54 untreated 36 thyroxine 12

I.Q.56Il “less gifted”

{ {

I.&.27( 1 “gifted” Over 5 years

1 1

I.Q.554 “less gifted’ I.Q.25E “gifted”

“Inflexion” mean & S.D. + 0.140k 0.754 + 0.454 2 0.709 - 0.062 2 0.720 - 0.201 + 0.724 + 0.005 & 0.307 -0.001 & 0.481 -0.189 k 0.442 - 0.061 t 0.389

Significance

N.S. P