Analytical and Clinical Validity Study of FirstStepDx PLUS: A Chromosomal Microarray Optimized for Patients with Neurodevelopmental Conditions.

INTRODUCTION: Chromosomal microarray analysis (CMA) is recognized as the first-tier test in the genetic evaluation of children with developmental delays, intellectual disabilities, congenital anomalies and autism spectrum disorders of unknown etiology. ARRAY DESIGN: To optimize detection of clinically relevant copy number variants associated with these conditions, we designed a whole-genome microarray, FirstStepDx PLUS (FSDX). A set of 88,435 custom probes was added to the Affymetrix CytoScanHD platform targeting genomic regions strongly associated with these conditions. This combination of 2,784,985 total probes results in the highest probe coverage and clinical yield for these disorders. RESULTS AND DISCUSSION: Clinical testing of this patient population is validated on DNA from either non-invasive buccal swabs or traditional blood samples. In this report we provide data demonstrating the analytic and clinical validity of FSDX and provide an overview of results from the first 7,570 consecutive patients tested clinically. We further demonstrate that buccal sampling is an effective method of obtaining DNA samples, which may provide improved results compared to traditional blood sampling for patients with neurodevelopmental disorders who exhibit somatic mosaicism.

SCN8A mutation in a child presenting with seizures and developmental delays.

The SCN8A gene encodes the sodium voltage-gated channel alpha subunit 8. Mutations in this gene have been associated with early infantile epileptic encephalopathy type 13. With the use of whole-exome sequencing, a de novo missense mutation in SCN8A was identified in a 4-yr-old female who initially exhibited symptoms of epilepsy at the age of 5 mo that progressed to a severe condition with very little movement, including being unable to sit or walk on her own.

KBG syndrome involving a single-nucleotide duplication in ANKRD11.

KBG syndrome is a rare autosomal dominant genetic condition characterized by neurological involvement and distinct facial, hand, and skeletal features. More than 70 cases have been reported; however, it is likely that KBG syndrome is underdiagnosed because of lack of comprehensive characterization of the heterogeneous phenotypic features. We describe the clinical manifestations in a male currently 13 years of age, who exhibited symptoms including epilepsy, severe developmental delay, distinct facial features, and hand anomalies, without a positive genetic diagnosis. Subsequent exome sequencing identified a novel de novo heterozygous single base pair duplication (c.6015dupA) in ANKRD11, which was validated by Sanger sequencing. This single-nucleotide duplication is predicted to lead to a premature stop codon and loss of function in ANKRD11, thereby implicating it as contributing to the proband's symptoms and yielding a molecular diagnosis of KBG syndrome. Before molecular diagnosis, this syndrome was not recognized in the proband, as several key features of the disorder were mild and were not recognized by clinicians, further supporting the concept of variable expressivity in many disorders. Although a diagnosis of cerebral folate deficiency has also been given, its significance for the proband's condition remains uncertain.

Haplotypes at LBX1 have distinct inheritance patterns with opposite effects in adolescent idiopathic scoliosis.

Adolescent idiopathic scoliosis (AIS) is a clinically significant disorder with high heritability that affects 2-4% of the population. Genome-wide association studies have identified LBX1 as a strong susceptibility locus for AIS in Asian and Caucasian populations. Here we further dissect the genetic association with AIS in a Caucasian population. To identify genetic markers associated with AIS we employed a genome-wide association study (GWAS) design comparing 620 female Caucasian patients who developed idiopathic scoliosis during adolescence with 1,287 ethnically matched females who had normal spinal curves by skeletal maturity. The genomic region around LBX1 was imputed and haplotypes investigated for genetic signals under different inheritance models. The strongest signal was identified upstream of LBX1 (rs11190878, P(trend) = 4.18 × 10(-9), OR = 0.63[0.54-0.74]). None of the remaining SNPs pass the genome-wide significance threshold. We found rs11190870, downstream of LBX1 and previously associated with AIS in Asian populations, to be in modest linkage disequilibrium (LD) with rs11190878 (r(2) = 0.40, D' = 0.81). Haplotype analysis shows that rs11190870 and rs11190878 track a single risk factor that resides on the ancestral haplotype and is shared across ethnic groups. We identify six haplotypes at the LBX1 locus including two strongly associated haplotypes; a recessive risk haplotype (TTA, Control(freq) = 0.52, P = 1.25 × 10(-9), OR = 1.56), and a co-dominant protective haplotype (CCG, Control(freq) = 0.28, P = 2.75 × 10(-7), OR = 0.65). Together the association signals from LBX1 explain 1.4% of phenotypic variance. Our results identify two clinically relevant haplotypes in the LBX1-region with opposite effects on AIS risk. The study demonstrates the utility of haplotypes over un-phased SNPs for individualized risk assessment by more strongly delineating individuals at risk for AIS without compromising the effect size.

FOXP3 gene polymorphisms in preeclampsia.

OBJECTIVE: To determine whether polymorphisms in the FOXP3 gene are associated with preeclampsia. STUDY DESIGN: Case-control study in which 120 women with preeclampsia were compared with 120 healthy normotensive controls. Genetic variants (single nucleotide polymorphisms and microsatellites) in the FOXP3 gene were analyzed. Polymorphisms were chosen based on studies of the FOXP3 gene in other autoimmune disorders. Correction of P values for multiple comparisons was performed by using the Benjamini-Hochberg procedure. RESULTS: There were no differences in the genotypes or allele frequencies in the single nucleotide polymorphisms between cases and controls. The FOXP3 GT microsatellite allele at 266 bp was less common in cases than controls (1.0% vs 5.2%, P = .0264). However, this did not remain significant after correction for multiple comparisons. CONCLUSION: Preeclampsia is not associated with FOXP3 gene polymorphisms that have been associated with other autoimmune disorders.

Vascular endothelial growth factor-A gene polymorphisms in women with recurrent pregnancy loss.

Vascular endothelial growth factor-A (VEGFA) is normally expressed at high levels in the human placenta, and lower levels have been observed in placental tissue of women with recurrent pregnancy loss. The objective of this study was to determine if genetic polymorphisms in the VEGFA gene associated with altered gene expression play a role in some cases of recurrent pregnancy loss (RPL). A case-control study of 99 women with RPL and 181 fertile controls was performed evaluating four common VEGFA polymorphisms associated with altered gene expression (-2578 C/A, -1154 G/A, -634 G/C, and +936 C/T). The allele frequency of the -2578 A allele was lower among women with RPL compared to fertile controls (0.39 vs. 0.48, p=0.049), while the allele frequency of the -634 C allele was higher among women with RPL compared to fertile controls (0.39 vs. 0.29, p=0.020). Women with RPL and controls had similar allele frequencies for the -1154 and +936 minor alleles. We conclude that some allelic polymorphisms associated with altered expression of VEGFA are more common among women with RPL compared to fertile controls.

Genetic variants in melatonin synthesis and signaling pathway are not associated with adolescent idiopathic scoliosis.

STUDY DESIGN: Genetic association study investigating the association of genetic markers of melatonin signaling and biosynthesis with adolescent idiopathic scoliosis (AIS). OBJECTIVE: To determine whether gene polymorphisms related to the melatonin signaling or biosynthesis pathways are associated with AIS. SUMMARY OF BACKGROUND DATA: Data have been published on the potential role of gene polymorphisms for melatonin receptor (MTNR) 1B in predicting AIS. Other genes in the melatonin pathways have been tested for association with AIS. METHODS: The following genes involved in melatonin synthesis were evaluated herein: tryptophan 5-hyroxylase 1 (TPH1), serotonin N-acetyltransferase (SNAT), and hydroxyindoleo-methyltransferase (HIOMT). In addition, proteins involved in melatonin signaling were also included in this study: MTNR1A, MTNR1B, and protein kinase C delta (PKCd). High throughput microarray-based single nucleotide polymorphism (SNP) genotyping was performed for these seven genes using DNA samples from 589 AIS subjects and 1533 ethnically matched controls. Chi-square analyses of allele frequency between AIS cases and controls were performed and odds ratios were calculated for all SNP markers. RESULTS: Three SNPs were tested for both MTNR1A and HIOMT, 4 for TPH1 and SNAT, 12 for PKCd, and 7 for MTNR1B. The minor allele frequencies were not significantly different between AIS cases and controls. No association was thus found between AIS and the investigated SNPs. CONCLUSIONS: Genetic polymorphisms associated with either melatonin synthesis or its signaling pathway are unlikely to be commonly associated with AIS.

Validation of DNA-based prognostic testing to predict spinal curve progression in adolescent idiopathic scoliosis.

STUDY DESIGN: Validation of a prognostic DNA marker panel. OBJECTIVE: The goals of this study were to develop and test the negative predictive value of a prognostic DNA test for adolescent idiopathic scoliosis (AIS) and to establish clinically meaningful endpoints for the test. SUMMARY OF BACKGROUND DATA: Clinical features do not adequately predict which children diagnosed with minimal or mild AIS will have the progressive form of the disease; genetic markers associated with curve progression might offer clinically useful prognostic insights. METHODS: Logistic regression was used to develop an algorithm to predict spinal curve progression incorporating genotypes for 53 single nucleotide polymorphisms and the patient's presenting spinal curve (Cobb angle). Three cohorts with known AIS outcomes were selected to reflect intended-use populations with various rates of AIS progression: 277 low-risk females representing a screening cohort, 257 females representing higher risk patients followed at referral centers, and 163 high risk males. DNA was extracted from saliva, and genotypes were determined using TaqMan assays. AIS Prognostic Test scores ranging from 1 to 200 were calculated. RESULTS: Low-risk scores (<41) had negative predictive values of 100%, 99%, and 97%, respectively, in the tested populations. In the risk model, we used cutoff scores of 50 and 180 to identify 75% of patients as low-risk (<1% risk of progressing to a surgical curve), 24% as intermediate-risk, and 1% as high-risk. CONCLUSION: Prognostic testing for AIS has the potential to reduce psychological trauma, serial exposure to diagnostic radiation, unnecessary treatments, and direct and indirect costs-of-care related to scoliosis monitoring in low-risk patients. Further improvements in test performance are expected as the optimal markers for each locus are identified and the underlying biologic pathways are better understood. The validity of the test applies only to white AIS patients; versions of the test optimized for AIS patients of other races have yet to be developed.

Polygenic inheritance of adolescent idiopathic scoliosis: a study of extended families in Utah.

A heritability study of 69 extended Utah families with a history of adolescent idiopathic scoliosis (AIS) indicates that AIS is a polygenic, multifactorial condition. Each family reported a history of AIS within four generations; a total of 247 individuals were confirmed via X-rays and medical records to have AIS. Coefficient of kinship was more than 25 standard deviations higher for these 69 families than for the general population. Excluding all probands and assuming autosomal dominant inheritance, 1,260 individuals over the age of 16 were determined to be at risk for AIS because they have a parent with AIS. Assuming 50% of these individuals carry the allele, estimated penetrance in at-risk males is approximately 9%, and estimated penetrance in at-risk females is approximately 29%. Recurrence risk in relatives decreases as the degree of relationship to the affected individual becomes more distant; however, the lowest recurrence risk calculated, for third-degree relatives, is still an average of 9%, well above the general population's risk. Onset of AIS appears to be inherited separate from curve pattern and severity. In a study of phenotypes in 36 of the families, the affected individuals were consistent in either curve severity or curve pattern, but not both. It is unclear whether severity or pattern is more heritable, but it is possible that the location of the curve on the spine is the most heritable trait of the phenotype. The study demonstrates the genetic complexity of AIS, including the low penetrance of its cumulative alleles and variable expression.

The association of innate immune response gene polymorphisms and puerperal group A streptococcal sepsis.

OBJECTIVE: The objective of the study was to determine whether single-nucleotide polymorphisms (SNPs) that influence the maternal innate immune response are associated with puerperal group A streptococcal sepsis. STUDY DESIGN: Subjects with confirmed puerperal group A streptococal infection were prospectively identified in 2 tertiary care hospitals over 18 years. Controls were racially matched subjects with term, uncomplicated deliveries. Thirty-eight polymorphisms associated with the innate immune response to bacterial infection were analyzed. Allele and genotype frequencies for subjects and controls were compared. RESULTS: Forty-eight women with puerperal group A streptococcal infection were identified. DNA was obtained for 28 subjects and 54 controls. Allele frequencies were significantly different between subjects and controls for polymorphisms in Toll-like receptor (TLR) 9-1486 (P = .03) and heat shock protein (HSP) 70-2 1267 (P = .003). Genotype frequencies were significantly different between subjects and controls for TLR9-1486 (P = .025), HSP70-2 1267 (P = .02), and interleukin (IL)-1beta-511 (P = .016). CONCLUSION: Puerperal group A streptococcal sepsis may be associated with innate immune response gene polymorphisms in TLR9, HSP70-2, and IL1beta.

Progesterone receptor genotype, family history, and spontaneous preterm birth.

OBJECTIVE: To examine whether women with a personal or family history of preterm birth are more likely to have genetic variation in the human progesterone receptor (hPR). METHODS: Women with a singleton preterm birth at less than 37 weeks of gestation between 2002 and 2006 were identified from a prospectively collected clinical and biologic obstetric database (cases). Women in the control group were those with only term deliveries at or above 38 weeks of gestation. The Utah Population Database was queried for family history (first- or second-degree relative) of preterm birth. DNA was extracted from stored buffy coats and genotyped for six single nucleotide polymorphisms in the hPR. RESULTS: One hundred fifty-four patients (92 women in the preterm case group, 62 women in the term control group) were included. All were white or Hispanic. There were no statistical differences between white and Hispanic allele frequencies. Women in the preterm case group were more likely to carry the minor allele, G (minor allele frequency 0.29 compared with 0.18, P=.035) for rs471767, and were more likely to carry the GT haplotype across rs471767 and rs578029 compared with women in the term control group. Similar haplotype block variation was seen among women with preterm birth plus a family history of preterm birth. CONCLUSION: Allele and haplotype frequencies in the hPR are significantly different among women with preterm birth and women with preterm birth plus a family history of preterm birth. This suggests the hPR gene may be a candidate for association with preterm birth or familial preterm birth. LEVEL OF EVIDENCE: III.

Polymorphisms in the promoter region of the interleukin-10 (IL-10) gene in women with cervical insufficiency.

OBJECTIVE: Our objective was to determine whether polymorphisms in the promoter region of the interleukin-10 gene are more common in women with cervical insufficiency compared with controls. STUDY DESIGN: We conducted a case-control study. Histories and blood were obtained from 121 cases and 157 controls. DNA was extracted and purified by using Puregene kits. Samples were analyzed for the interleukin-10 -1082 G/A polymorphism and interleukin-10.G microsatellite in the interleukin-10 gene promoter region. RESULTS: The -1082 G/A polymorphism in the promoter region of the interleukin-10 gene occurred with similar frequency in cases and controls. The interleukin-10.G microsatellite contained 10 alleles (G6-G16). The G13 allele was present more frequently in cases (24.1%) compared with controls (14.6%) (P = .05). CONCLUSION: Interleukin-10 is known to down-regulate inflammation. The G13 allele in the interleukin-10.G microsatellite occurred more frequently in women with cervical insufficiency compared with controls, suggesting that alterations in inflammatory processes may play a role in cervical insufficiency.

Simultaneous amplification, detection, and analysis of common mutations in the galactose-1-phosphate uridyl transferase gene.

Classic galactosemia is an autosomal recessive inherited error of galactose metabolism. It is caused by lack of galactose-1-phosphate uridyl transferase, an enzyme that is required to metabolize galactose-1-phosphate to uridine diphosphate galactose. The build up of galactose-1-phosphate is toxic at high levels and can damage the liver, brain, eyes, and other vital organs. Over 200 mutations have been identified in affected individuals. We describe an assay to identify nine target mutations or variants in the galactose-1-phosphate uridyl transferase gene, namely p.Q188R, p.S135L, p.K285N, p.L195P, p.T138M, p.Y209C, IVS2-2 A>G, p.L218L, and p.N314D. A single long-range PCR is followed by a multiplexed nucleotide extension assay (single nucleotide extension) and capillary electrophoresis to detect simultaneously all nine target mutations/variants. Fifty-four previously characterized samples (47 clinical samples and seven controls) gave a 100% concordance. We also report a nontarget novel mutation, p.L192X, and its profile using single nucleotide extension. This assay can complement the enzyme activity assay and identify familial mutations for testing additional family members.

Collagen 1Alpha1 and transforming growth factor-beta polymorphisms in women with cervical insufficiency.

OBJECTIVE: To estimate whether polymorphisms in the collagen 1Alpha1 gene (COL1Alpha1) and the transforming growth factor-beta gene (TGF-beta;1) are more common in women with cervical insufficiency than in those without the condition. METHODS: Medical, obstetric, and family histories and blood were obtained from women with (n=121) and those without (n=165) cervical insufficiency. DNA was extracted and purified by using commercial DNA isolation kits. Samples were analyzed for variants in two genes, the COL1A1 intron 1SP1 and TGF-beta Arg-25-Pro polymorphism, by using an allele-specific polymerase chain reaction assay. RESULTS: Thirty-four of 125 (27.2%) women with cervical insufficiency had at least one first-degree female relative affected. The frequency of the homozygous TT genotype in the COL1A1 gene was increased in women with a history of cervical insufficiency compared with controls (10.8% compared with 3.1%, P=.04). The TGF-beta polymorphisms (ArgPro and ProPro) also were increased in cases (38.3% compared with 14.6%, P<.001). CONCLUSION: Over one fourth of women with cervical insufficiency have a family history of cervical insufficiency, and the COL1A1 intron 1SP1 and TGF-beta Arg-25-Pro polymorphisms are associated with the condition. These observations suggest that, in part, cervical insufficiency is mediated by genetic factors. LEVEL OF EVIDENCE: II.

The search for idiopathic scoliosis genes.

STUDY DESIGN: A cohort of 145 patients with adolescent idiopathic scoliosis (AIS) were identified and contacted to determine whether they had a family history of scoliosis. These results were submitted to an internal genealogical database to screen for potential connections to other AIS families. The severity and incidence of AIS in extended family groups were also analyzed. OBJECTIVES: Our objectives were to quantify the genetic effect in AIS, determine the expressivity and penetrance of AIS in large family groupings, and examine larger scoliosis pedigrees for evidence of multiple genes. SUMMARY OF BACKGROUND DATA: Previous reports have suggested an 80% connectedness among scoliosis families, but no clear evidence of multiple genes. It is not known if there are major gene(s). METHODS: A cohort of 145 AIS probands were identified and contacted to ascertain whether they had a family history of AIS. Their medical records and spine radiographs were reviewed to confirm the diagnosis and determine the disease severity. Using an internal genealogical database, the cases were screened for potential connections that would produce larger extended pedigrees. RESULTS: Overall, 131 of the probands were in the database and 127 showed connections to other scoliosis families, a 97% connectedness. These results suggest a major scoliosis gene, as more than 50% of the probands were connected by founders that all resided in England in the mid 1500s. The differences in penetrance (41% vs. 34%) and expressivity (38% vs. 61%) between seemingly unrelated large family groupings might suggest that two different genes are a major influence for AIS in these families. CONCLUSIONS: Nearly all (97%) AIS patients have familial origins. There appears to be at least one major gene, and the differences in penetrance and expressivity in two large unconnected pedigrees might suggest the presence of more than one gene.

Genotyping of human platelet antigens 1 to 6 and 15 by high-resolution amplicon melting and conventional hybridization probes.

High-resolution melting techniques are a simple and cost-effective alternative to other closed-tube genotyping methods. Here, we genotyped human platelet antigens (HPAs) 1 to 6 and 15 by high-resolution melting methods that did not require labeled probes. Conventional melting analysis with hybridization probes (HybProbes) was also performed at each locus. HybProbe assays were performed individually, whereas amplicon melting (HPAs 1 to 5 and 16) and unlabeled probe (HPA 6) assays were duplexed when possible. At all loci for each method, both homozygous and heterozygous genotypes were easily identified. We analyzed 100 blinded clinical samples (33 amniotic fluid, 12 cultured amniocytes, and 55 blood samples) for all 7 single-nucleotide polymorphisms (SNPs) by each method. Genotype assignments could be made in 99.0% of the SNPs by high-resolution melting and in 98.7% of the SNPs with HybProbes with an overall genotype concordance of 98.8%. Errors included two sample misidentifications and six incorrect assignments that were all resolved by repeating the analysis. Advantages of high-resolution melting include rapid assay development and execution, no need for modified oligonucleotides, and similar accuracy in genotyping compared with other closed-tube melting methods.

The heritability of preterm delivery.

OBJECTIVE: To study the heritability of preterm delivery. METHODS: Women who delivered a singleton infant at less than 36 weeks of gestation were asked about their family history. Twenty-eight families were identified in which the proband had at least five first- or second-degree relatives with preterm delivery. An extensive genealogy database (GenDB) was constructed using more than 9,000 genealogy sources in the public domain (records before 1929). GenDB documents the relationships between more than 17.5 million ancestors and 3.5 million descendants of approximately 10,000 individuals who moved to Utah in the mid 1800s. This database was searched for the names, birth dates, and birthplaces of the four grandparents for each of the 28 probands. Pairwise coefficients of kinship were determined for the 93 preterm delivery grandparents identified, and for sets of 100 individuals born in the 1920s who were randomly selected from the population database. RESULTS: Probands had a mean of 3.3 grandparents included in this database. The average coefficient of kinship for controls was 1.5 x 10(6) (standard deviation = 0.6 x 10(6)). This measure agrees with previous calculations for the Utah population. The coefficient of kinship for familial preterm delivery grandparents was more than 50 standard deviations higher (3.4 x 10(5) [P < .001]). CONCLUSION: This study confirms the familial nature of preterm delivery. On average, gravidae randomly selected from our population are 23rd degree relatives, while these preterm delivery probands are eighth-degree relatives. A genome-wide scan using these affected families is underway.

The factor V Leiden and the G20210A prothrombin gene mutations are rare in women with fetal death.

PROBLEM: To determine if there is an association between two commonly inherited thrombophilias, the factor V Leiden and the G20210A prothrombin mutations, and fetal death. METHOD OF STUDY: We used a case-control study design to compare the frequencies of these mutations in women with fetal death and controls. Fetal death was the intrauterine death of the conceptus > or =10 weeks gestation. Controls had one live birth, no miscarriages, and no fetal death. Results were compared using chi square analysis. RESULTS: One hundred and seventy-five cases and controls were identified. There were 4.6% of cases and 3.8% of controls heterozygous for the factor V Leiden mutation (NS), and 1.3% of cases and 1.7% of controls heterozygous for the prothrombin mutation (NS). CONCLUSION: In our population, neither the factor V Leiden nor the G20210A prothrombin mutations are associated with fetal death. Further evidence is required before routine screening for these mutations can be recommended.

Homocysteine levels in women with antiphospholipid syndrome and normal fertile controls.

OBJECTIVE: Recent studies have identified hyperhomocysteinemia as a risk factor for both recurrent pregnancy loss and thrombosis. Antiphospholipid syndrome, an autoimmune disorder, is also characterized by recurrent pregnancy loss and thrombosis. Thus, our purpose was to determine if hyperhomocysteinemia is more common in patients with APS than normal fertile controls. METHODS: Plasma, sera and whole blood were obtained from two groups of women: (1) 22 with well-characterized antiphospholipid syndrome; and (2) 41 healthy fertile controls. Levels of fasting homocysteine, vitamin B16, vitamin B12, folate and the incidence of the C677/T mutation of the methylene tetrahydrofolate reductase genotype (C677T/MTHFR) were determined. RESULTS: The proportion of individuals with hyperhomocysteinemia and fasting plasma homocysteine levels were similar in women with APS and controls. Levels of vitamin B6, vitamin B12, folate and the incidence of C677/MTHFR were also similar in the two groups. CONCLUSION: Hyperhomocyteinemia and the C677T/MTHFR mutation are not common in women with antiphospholipid syndrome. Abnormal homocysteine metabolism is unlikely to play a major role in the pathogenesis of antiphospholipid syndrome.

Prenatal diagnosis of 21-hydroxylase deficiency caused by gene conversion and rearrangements: pitfalls and molecular diagnostic solutions.

OBJECTIVES: The present paper reports the prenatal diagnosis of congenital adrenal hyperplasia (CAH) in two cases of 21-hydroxylase deficiency. DNA diagnostic errors can be caused by the presence of the highly homologous 21-hydroxylase pseudogene, CYP21P, adjacent to the functional gene, CYP21. The present paper details how complex gene conversions and rearrangements between the CYP21 and CYP21P pose unique complications for prenatal diagnosis. METHODS: Analysis of eight common mutations in the 21-hydroxylase gene as well as deletion of the entire gene is accomplished using polymerase chin reaction (PCR) followed by amplified created restriction site (ACRS) or allele-specific oligohybridization (ASO) and Southern blot followed by hybridization to a CYP21-specific probe. Linkage analysis was performed using microsatellite markers flanking the CYP21 gene. RESULTS: The direct mutation detection assay indicated a complicated gene conversion and rearrangement in the probands of both families. Interpretation of these rearrangements made it difficult to determine whether or not the fetuses would be affected with CAH. Linkage studies revealed that each fetus had inherited both parental disease chromosomes and was therefore predicted to be affected with CAH. CONCLUSION: As observed in the two reported cases, direct DNA analysis may provide limited information due to gene conversion or rearrangement between the CYP21 and CYP21P genes. These cases suggest that direct mutation detection should be supported by linkage analysis, whenever possible, to provide more comprehensive information for the family.