Molecular analysis of the MVK and TNFRSF1A genes in patients with a clinical presentation typical of the hyperimmunoglobulinemia D with periodic fever syndrome: A low-penetrance TNFRSF1A variant in a heterozygous MVK carrier possibly influences the phenotype of hyperimmunoglobulinemia D with periodic fever syndrome or vice versa




To describe biochemical findings and the spectrum of mevalonate kinase (MVK) gene mutations as well as an associated TNFRSF1A low-penetrance variant in a series of patients with clinical features of the hyperimmunoglobulinemia D with periodic fever syndrome (HIDS).


The MVK gene was sequenced in 8 children and 1 adult (including 2 siblings) fulfilling the clinical criteria for HIDS. In addition, sequencing of exons 2, 3, 4, and 6 of the TNFRSF1A gene was performed in patients with only one or no MVK mutation. Mevalonate kinase (MK) enzyme activity in leukocytes and renal excretion of mevalonic acid were also measured.


Mutations in the coding region of the MVK gene were detected in 6 patients, and the most common mutation was V377I. Among these patients were 2 novel mutations, both of which were located in exon 6. These novel mutations resulted in the substitution of tryptophan (TGG) by a stop codon (TGA) at amino acid position 188 (W188X) and in the exchange of valine (GTG) for alanine (GCG) at amino acid position 203 (V203A). In 1 patient, a combination of one MVK (V377I) mutation and one TNFRSF1A (R92Q) mutation was present. The patient's clinical phenotype resembled a mixture of variant-type HIDS and tumor necrosis factor receptor–associated periodic syndrome (TRAPS). Her IgD values varied between normal and slightly increased, and the MK activity was in the low-normal range, while urinary mevalonate concentrations were always normal.


The genotype findings indicate that a relatively small number of genes may be involved in the clinical manifestation of HIDS, with low-penetrance TNFRSF1A variants possibly influencing the HIDS phenotype or MVK mutations contributing to TRAPS.

The hyperimmunoglobulinemia D with periodic fever syndrome (HIDS; MIM no. 260920) is an autosomal recessively inherited autoinflammatory disease. It is characterized by febrile episodes of 3–7 days' duration recurring every 4–8 weeks and by a persistently high serum level of IgD (>100 IU/ml). Symptoms accompanying a typical attack comprise cervical lymphadenopathy, chills, headache, abdominal pain, vomiting, diarrhea, arthralgia, arthritis, skin rash, splenomegaly, hepatomegaly, and/or oral ulcers. Another concomitant finding is a strong acute-phase reaction with leukocytosis, elevated C-reactive protein levels, and a high erythrocyte sedimentation rate. The prognosis is good, since there is no development of amyloidosis (1–3).

HIDS is caused by mutations in the mevalonate kinase (MVK) gene on chromosome 12q24 which lead to a depressed enzymatic activity of mevalonate kinase (MK) (4, 5). MK is a key enzyme of cholesterol and isoprenoid biosynthesis and is present in the cytosol and peroxisomes of every mammalian cell. At an early step of cholesterol synthesis, it phosphorylates 3,R-mevalonic acid to 5-phosphomevalonate. In HIDS patients, MK activity in cultured skin fibroblasts and lymphocytes is reduced to 1–12% compared with that in controls (4–6). This leads to a slightly increased urinary excretion of mevalonic acid (<20 mmoles/mole of creatinine) during a febrile episode. In contrast, in the case of a very profound MK deficiency, usually with <0.5% of the enzymatic activity found in controls, patients present with the clinical phenotype of mevalonic aciduria (MA; MIM no. 251170), and renal excretion of mevalonic acid is much higher (from ∼3,000 mmoles/mole of creatinine up to >56,000 mmoles/mole of creatinine). In addition to the typical clinical picture of HIDS, MA patients present with psychomotor retardation, facial dysmorphia, cataract, failure to thrive, and ataxia (6–8). Thus, MK deficiency comprises a continuous spectrum of loss of enzymatic activity, with the severe phenotype of MA at the most pathologic extreme, occurring in <1% of MK-deficient patients, and with HIDS as the milder variant.

More than 30 mutations have been reported in all protein-coding exons of the MVK gene except exon 4, and the vast majority are missense mutations. The most common genetic defect is the substitution of valine by isoleucine at amino acid position 377 (V377I), which is found in >90% of HIDS patients. The second mutation is often one that has also been associated with MA, which suggests that it results in a nonfunctional enzyme. This indicates that the V377I substitution is responsible for the HIDS phenotype.

In addition, patients have been described with a clinical presentation typical of HIDS, but without MVK mutations. This HIDS subgroup, termed variant-type HIDS, is characterized by normal or only slightly decreased MK activities (9, 10). Given the phenotypic similarities, variant-type HIDS raises two important questions: 1) are there HIDS-related mutations in genes other than MVK? and 2) what are the pathophysiologic mechanisms involved (e.g., defective isoprenylation of key proteins [11])?

Tumor necrosis factor receptor–associated periodic syndrome (TRAPS; MIM no. 142680), on the other hand, is the most frequent autosomal dominantly inherited periodic fever syndrome. This condition is caused by mutations in exons 2–4 and 6 of the TNFRSF1A gene, which codes for the extracellular, soluble part of TNF receptor superfamily 1A (TNFRSF1A). Most common are the low-penetrance mutations P46L and R92Q, which were also present on ∼1% of healthy control chromosomes in one study (12). One HIDS patient with 2 MVK mutations has been reported to carry in addition the TNFRSF1A P46L variant on 1 allele (13).

Here, we report on the spectrum of MVK mutations in a group of HIDS patients with variable MK activities. We present the case of a proband with 1 MVK mutation and 1 low-penetrance TNFRSF1A mutation who exhibited symptoms of HIDS and TRAPS.


The study was approved by the Ethics Committee of the Medical Faculty, Ludwig-Maximilians-University, Munich, Germany, and was performed according to the most recent version of the Declaration of Helsinki. Written informed consent was obtained from all subjects or their parents prior to enrollment.


Probands are identified by a capital letter and, in the case of members from one family, by a capital letter in combination with a number. Patients' charts and laboratory results (e.g., acute-phase reactant and immunoglobulin levels) were reviewed, and the probands and/or their parents were interviewed with regard to clinical symptoms and disease history. For participation in this study, patients were selected according to the following clinical criteria for HIDS: 1) recurrent febrile episodes with a temperature >38.5°C not explained by infections or otherwise and 2) serum IgD concentrations >100 IU/ml measured at least 2 times with a minimum interval of 1 month. If available, parents of mutation-positive probands were also analyzed.

Analysis of leukocyte MK activity and urinary mevalonate excretion.

MK activity was measured in leukocytes prepared from 2–3 ml of peripheral blood treated with EDTA. The assay was performed as previously described (14). 14C-mevalonate-5-phosphate was separated from the substrate 14C-mevalonic acid on a mini DEAE-cellulose column (15). The normal range is 0.4–1.0 nmoles/minute/mg of protein.

Urine samples were collected during and between febrile attacks. Mevalonate excretion was determined by gas chromatography mass spectrometry (GC-MS) using a modification of a previously described stable isotope dilution assay (16). Prior to analysis, urine samples were acidified with 130 μl 6N HCl to convert mevalonic acid to mevalonolactone. Twenty-five microliters of 1 mM D3-mevalonolactone (C/D/N Isotopes, Pointe Claire, Quebec, Canada) was added as internal standard. Mevalonolactone was extracted by liquid–liquid partition using Extrelut NT (Merck, Darmstadt, Germany). A mixture of CHCl3 and 2-methyl-2-butanol (95/5 volume/volume) was used as organic solvent. The solvent was evaporated under a stream of nitrogen, and the residue was derivatized with 50 μl N-methyl-N-trimethylsilylheptafluorobutyramide for 1 hour at 60°C to form the trimethylsilyl derivative of mevalonolactone.

For GC-MS analysis, the MSD 5972 quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA) was run in the selective ion monitoring mode (mass-to-charge ratio [m/z] 145 for mevalonolactone and m/z 148 for the deuterated standard). One microliter of derivatized sample was injected into the column in splitless mode. Gas chromatographic separation was achieved on a DB-5MS capillary column (30 meters × 0.25 mm, 0.25 μm; J&W Scientific, Agilent Technologies) using helium as carrier gas. The initial temperature was 80°C for 5 minutes, and this was increased to 280°C at a rate of 4°C per minute.

DNA and RNA extraction.

Genomic DNA was isolated from nucleated blood cells of the probands and, if available, from those of their parents with the help of the QIAamp blood mini kit (Qiagen, Hilden, Germany). Total RNA was purified from peripheral lymphocytes and monocytes obtained from patient F as well as from 11 controls by using the TRIzol reagent (Invitrogen, Carlsbad, CA).

Reverse transcription of total RNA.

Full-length and partial complementary copies of MVK messenger RNA were obtained by first-strand complementary DNA (cDNA) synthesis of 1 μg of total RNA in a 10-μl reaction volume with Superscript II RNase H reverse transcriptase (Invitrogen) using random hexamer primers (Roche Diagnostics, Mannheim, Germany).

Complementary DNA and DNA amplification by polymerase chain reaction (PCR).

The MVK cDNA was amplified from one-half of the reverse transcription reaction, and the 11 exons of the MVK gene as well as exons 2, 3, 4, and 6 of the TNFRSF1A gene were in each case amplified from ∼500 ng of genomic DNA with 1.25 units HotStar Taq DNA polymerase (Qiagen) in a 50-μl reaction volume containing the PCR buffer supplied by the manufacturer, 1.5 mM MgCl2, and, in the case of MVK exon 1, Q-solution (Qiagen). The cycling profile consisted of an initial 15-minute enzyme activation step at 95°C, followed by 40 cycles of denaturation at 95°C for 30 seconds, primer annealing at 62°C for 30 seconds, and cDNA synthesis at 72°C for 30 seconds.

MVK and TNFRSF1A mutation analysis.

Cleanup of exon amplification products was performed with the help of the QIAquick PCR purification kit (Qiagen), and recovered fragments were sequenced with the ABI Prism BigDye Terminator v3.1 Ready Reaction Cycle Sequencing kit (Applied Biosystems, Foster City, CA) following the procedures recommended by the manufacturers. Sequences were analyzed on an ABI Prism 377 DNA Sequencer (Applied Biosystems) using the Sequence Analysis program version 3.4.5 (Applied Biosystems).


The main clinical signs/symptoms of the 9 HIDS patients who participated in this study are summarized in Table 1. The male:female ratio was 4:5. Patients C1 and C2 were sisters. All other patients were unrelated Caucasians of German ancestry, except for proband H, who was of German African origin, and patient E, who was of German English origin (Table 2).

Table 1. Characteristics of the 9 patients with the clinical phenotype of hyperimmunoglobulinemia D with periodic fever syndrome*
  • *

    CRP = C-reactive protein; ESR = erythrocyte sedimentation rate.

  • Highest documented value.

  • Measured during an attack.

Age, years1510162217711109
Age at onset, months30.51134861927
Duration of fever, mean days3.55.510.57.5924.52.58.5
Fever-free interval, mean weeks2465335.5611
Fever after immunization++++
Abdominal pain+++++
Aphthous ulcers++++
IgDmax, IU/ml2891,4501,4701,550704148580240197
IgAmax, mg/dl7904301,380773762263393165159
Leukocytesmax, ×109/liter33.623.617.234.213.816.329.111.213.3
CRP levelmax, mg/dl24.417.311.725.121.911.39.39.215
ESRmax, mm/hour90708611011183502597
Table 2. Summary of the mevalonate kinase (MK) enzyme activities measured in peripheral blood leukocytes, the highest documented values of urinary excretion of mevalonic acid measured during an attack, and the MVK and TNFRSF1A gene analyses performed in the 9 probands*
Patient/ethnicityMK enzyme activityUrinary mevalonic acidMVK mutationTNFRSF1A mutation, amino acid change
ExonNucleotide changeAmino acid change
  • *

    ND = not done.

  • Expressed as nmoles/minute/mg of protein (normal range 0.4–1.0).

  • Expressed as mmoles/mole of creatinine (control values: mean ± SD 0.33 ± 0.22, range 0.09–0.77).

A/northern German0.0551.24111129 G→AV377IND
   111129 G→AV377I 
B/southern German0.0248.426564 G→AW188XND
   111129 G→AV377I 
C1/northern German0.00251.566608 T→CV203AND
   9803 T→CI268T 
C2/northern German0.00117.136608 T→CV203AND
   9803 T→CI268T 
D/southern German0.01ND111129 G→AV377I
E/southern German English0.50.21111129 G→AV377IR92Q
F/southern German0.83ND
G/northern German0.870.27
H/northern German African0.870.27

Mutation analysis.

In 5 of the 8 unrelated patients (probands C1 and C2 were sisters), we detected a total of 4 different MVK missense mutations, the most common being V377I (Table 2). Patients D and E were heterozygotes for V377I. Proband A was found to be a homozygote for V377I, and 3 patients were compound heterozygotes for either V377I and a novel mutation (patient B) or I268T and a different novel mutation (siblings C1 and C2). Both novel mutations were located in exon 6 (Figure 1). Patient B was the heterozygous carrier of a G564→A substitution, replacing tryptophan (TGG) with an Opal termination signal (TGA) at amino acid position 188 (W188X). Sisters C1 and C2, on the other hand, had inherited a T608→C mutation on 1 allele, leading to the exchange of valine (GTG) for alanine (GCG) at amino acid position 203 (V203A). Sequence analysis of the asymptomatic parents of the 2 siblings revealed a heterozygous carrier state for V203A in the mother and for I268T in the father. A search assisted by software (PC/GENE; IntelliGenetics, Mountain View, CA) demonstrated that the G564→A replacement results in the loss of two restriction sites for the enzymes Asu I/Sau96 I (5′-G/GNCC-3′) and Hae III (5′-GG/CC-3′). The T608→C substitution, on the other hand, creates new cleavage sites for the 3 enzymes Aci I (5′-C/CGC-3′), BsrB I (5′-CCG/CTC-3′), and Mwo I (5′-GCNNNNN/NNGC-3′).

Figure 1.

Novel MVK mutations W188X and V203A. Shown are sequence electropherograms of MVK exon 6 amplification products from 2 probands and 2 healthy controls. Arrows indicate the heterozygous TGG→TGA (W188X) substitution identified in patient B and the heterozygous GTG→GCG (V203A) substitution identified in patient C1. The latter substitution also appeared in patient C2, the sibling of patient C1.

In patients F, G, and H, sequence analysis failed to identify any changes in MVK exons 1–11 or TNFRSF1A exons 2, 3, 4, and 6 or in the adjacent intronic regions. These probands' clinical signs and symptoms (Table 1) were compatible with those of the variant-type HIDS patients described by Simon et al (9).

Polymorphisms/splice variants.

We detected 8 polymorphisms, of which 7 were located in the introns (Table 3). Molecular haplotyping revealed linkage disequilibrium between the −38 C→T exchange in intron 8 and the +24 G→A substitution in intron 9, while the C→G replacement in intron 1 at position −118 with regard to the acceptor splice site of exon 2 appeared not to be linked to the +61 A→G substitution at the 5′ end of intron 2. In addition, a G405→A transition was found within the coding sequence of exon 5, which does not result in an amino acid exchange (S135S).

Table 3. Sequence polymorphisms in the MVK gene identified in the present study*
IVSAcceptor/donor splice siteRelative locationSubstitution
  • *

    IVS = intervening sequence; as = acceptor splice site; ds = donor splice site.

Intron 1as−118C→G
Intron 2ds+61A→G
Intron 4ds+8C→T
Exon 5 S135STCG→TCA
Intron 6as−18A→G
Intron 8as−38C→T
Intron 8as−7T→G
Intron 9ds+24G→A

In search of causal genetic defects in proband F, who showed the clinical phenotype typical of HIDS but lacked known MVK mutations, we generated full-length and partial MVK cDNA fragments from total lymphocyte and monocyte RNA isolated from the patient and from 11 healthy controls in order to find out whether the intronic nucleotide substitutions found in this proband were polymorphisms or disease-causing mutations. This led to the detection of 5 different splice variants affecting primarily exons 4 and 5 of the MVK gene (Table 4). Three of these splice variants (deletion of exon 4, of exons 4 and 5, and of exons 4–7) result in the synthesis of 83–amino acid proteins that differ only in their last 8 carboxyl-terminal residues, which are encoded by MVK exons 5, 6, or 8 (Table 5).

Table 4. Naturally occurring splice variants of the MVK gene detected by reverse transcriptase–polymerase chain reaction of total lymphocyte and monocyte RNA isolated from 1 of the probands as well as from 11 normal controls*
Deleted exonConsequence
  • *

    MK = mevalonate kinase; aa = amino acid.

2, 4, and 5ATG at MK position 195; 202-aa protein
483-aa protein; C-terminal 8 aa encoded by exon 5
5In-frame fusion; 344-aa protein
4 and 583-aa protein; C-terminal 8 aa encoded by exon 6
4–783-aa protein; C-terminal 8 aa encoded by exon 8
Table 5. Sequence of the 8 novel carboxyl-terminal amino acids of the three 83-residue proteins that are the result of alternate splicing of the MVK mRNA*
Deleted exonC-terminal amino acid sequence
  • *

    Only the areas of difference are shown; the remainders of the proteins are identical.

4 and 5-G-P-C-R-A-W-I-S

MK enzyme activity and renal mevalonate excretion.

Results of the enzymatic assays in peripheral blood leukocytes and the urinary mevalonate concentrations of the study patients are given in Table 2. The 2 compound heterozygous sisters C1 and C2 had extremely low MK activities (close to zero), but presented with a phenotype typical of HIDS without MA-specific neurologic signs. Probands D and E, with at least 1 MVK mutation, also exhibited depressed MK activities of varying degrees. On the other hand, patients F, G, and H, with no detectable MVK mutation, had normal enzyme activities.

Urinary mevalonate concentrations during febrile episodes increased significantly in patients B, C1, and C2, each of whom carried 1 MA-associated mutation (8.42–51.56 mmoles/mole of creatinine) (Table 2); however, concentrations were also elevated between fever attacks, ranging from 2.9 mmoles/mole of creatinine to 13.77 mmoles/mole of creatinine (data not shown). While these results correlated with the depressed MK activity in these 3 probands, the V377I-homozygous patient A, who also presented with distinctly reduced enzymatic activity, had only slightly increased concentrations of urinary mevalonate even during episodes (Table 2).

V377I-heterozygous patients D and E.

Proband E, who was a compound heterozygote for V377I and the common R92Q mutation encoded by exon 4 of the TNFRSF1A gene, had presented with mild clinical features typical of HIDS (cervical lymphadenopathy, mild abdominal pain, arthralgia, splenomegaly) since the age of 4 years. IgD values varied between normal and slightly increased, and the MK activity was in the low-normal range (0.43–0.5 nmoles/minute/mg protein), while urinary mevalonate concentrations were always normal. Steroid treatment was highly effective in terms of prompt resolution of a febrile attack, which is typical of TRAPS.

Mutation analysis of her parents demonstrated that she had inherited the V377I mutation from her asymptomatic father, who was of English origin. Interestingly, his MK activity was slightly decreased (0.32 nmoles/minute/mg protein). The German mother of patient E was healthy and was a carrier of the R92Q TNFRSF1A variant. Her family history was negative for a periodic fever syndrome.

Patient D, however, was also a heterozygous carrier of the V377I mutation, but further analyses of the MVK and TNFRSF1A genes failed to reveal a second mutation. She had presented with a clinical phenotype typical of HIDS since the age of 3 months, had highly elevated IgD (>700 IU/ml) and IgA levels, and her MK activity was greatly diminished (0.01 nmoles/minute/mg protein).


Our study expands the spectrum of MVK mutations and polymorphisms, which are currently listed in the World Wide Web–accessible database INFEVERS (online at[17]). We also identified 2 previously undescribed, naturally occurring splice variants with combined deletions of exons 2, 4, and 5 as well as of exons 4–7.

The nonsense mutation W188X in patient B leads to premature termination of protein translation and to the synthesis of a truncated, and most likely inactive, enzyme. Together with heterozygosity for the common V377I mutation, this results in the classic HIDS phenotype with depressed MK enzyme activity. While we were in the process of preparing the manuscript for this report, this nonsense mutation was submitted to the INFEVERS database by F. T. Saulsbury in the form of a personal communication.

In the 2 sisters C1 and C2, we found the second most common I268T mutation in combination with the replacement of valine by alanine at amino acid position 203. The side chains of these 2 nonpolar amino acids are different, suggesting functional differences. In addition, the valine residue is conserved among mammalian species and is located in close proximity to Asp204, which is part of the catalytic site of the enzyme and functions as the catalytic base (18–20). In comparison with their mother, who was heterozygous for the V203A mutation and completely healthy, both sisters had nearly absent MK activities and strong febrile, HIDS-typical attacks. Therefore, the V203A substitution, in combination with the I268T mutation, appears to cause a severe form of HIDS.

As in other studies (5, 21, 22), V377I was the most frequent MVK mutation causing HIDS. Furthermore, also in accordance with earlier observations, the majority of patients were compound heterozygous for this genetic defect and another mutation. Interestingly, we also identified 2 patients who were solely heterozygous for this particular amino acid substitution. Proband E had also inherited the R92Q mutation, encoded by exon 4 of the TNFRSF1A gene, on the maternal allele. Accordingly, the patient exhibited a mixed phenotype, with symptoms of both disorders. She presented with mild clinical features typical of HIDS (cervical lymphadenopathy, mild abdominal pain, arthralgia, splenomegaly), had a low-normal MK activity with normal mevalonic acid excretion values, and responded very well to steroid therapy. This suggests that another MVK mutation was probably absent, which is supported by the results of the DNA sequence analyses. The patient's father was a healthy V377I heterozygote and showed a slightly depressed MK enzyme activity, while her mother was an asymptomatic carrier of the TNFRSF1A R92Q substitution.

Generally, R92Q is regarded as a low-penetrance mutation associated with a broader range of symptoms than most TRAPS mutations and found especially in sporadic cases of TRAPS (12), although segregation with disease has been observed in 4 families (23). The R92Q allele frequency ranges from 1.8% (23) to 3.3% (12) in patients with clinical symptoms suggestive of TRAPS and amounts to ∼1% in Irish and North American control populations (12). Although an in vitro shedding defect is absent, soluble TNFRSF1A levels did not increase during an inflammatory episode in an R92Q carrier, suggesting that this amino acid substitution leads to an in vivo dysfunctional receptor (12). Thus, one could also speculate that our patient had TRAPS with the V377I mutation probably contributing to the symptoms.

Arkwright et al (13) already described a patient who was a compound heterozygote for 2 MVK mutations (G211A and V377I) and who also carried the P46L TNFRSF1A variant on 1 allele. Subsequently, they tested 15 confirmed HIDS patients for the 3 low-penetrance mutations P46L, R92Q, and R92P. Since no additional TNFRSF1A mutations were detected, those investigators suggested that probands with HIDS have no increased frequency of TNFRSF1A mutations. However, based on our findings, we suggest that patients with periodic fever should be tested for TNFRSF1A mutations, irrespective of whether they carry 2, 1, or no MVK mutations. This could answer the question of whether HIDS patients have an increased frequency of TNFRSF1A mutations, which could contribute to the severity of the disease.

Patient D, carrying the V377I variant on only 1 allele, showed the typical clinical and laboratory features of HIDS, with high IgD and IgA levels and sharply decreased MK activity. However, we were unable to detect a second mutation in the MVK and TNFRSF1A genes of this patient. Unfortunately, the parents did not agree to a second withdrawal of blood from their daughter in order to confirm the low MK activity as well as to perform cDNA analyses. Blood was always drawn during the asymptomatic interval, since temperature is known to affect MK activity (24); thus, a false-positive result can be excluded in this case. The disease may therefore be the result of another alteration in the promoter region (of which only 35 bp directly upstream of the cap site were sequenced), in the 3′-untranslated region, or in intronic sequences of the MVK gene locus which were also not covered by our analyses.

In probands F, G, and H, with the clinical phenotype of HIDS, no MVK or TNFRSF1A mutation could be detected. Interestingly, these patients showed milder or no laboratory abnormalities, with normal MK activities (all 3 patients) and urinary mevalonic acid concentrations (tested for patients G and H only), only slightly elevated IgD levels (except for patient F), and normal IgA levels. Since other hereditary periodic fever syndromes have been excluded from a clinical and molecular genetic point of view, and since all patients displayed the typical HIDS phenotype, we consider these probands to be variant-type HIDS patients.

In summary, our study illustrates the clinical and genetic heterogeneity of HIDS. A possible association with low-penetrance TNFRSF1A mutations in heterozygous or variant-type HIDS patients needs to be investigated, since this is likely to influence the phenotype.


We thank the family members for participating in the study, Franz Jansen and Dieter Kunz for excellent technical assistance, Barbara Asam for secretarial assistance, and Dr. Claus-Dieter Langhans for establishing, and Patrik Feyh for performing, the urinary mevalonic acid measurements at Children's Hospital, University of Heidelberg, Heidelberg, Germany.