Circulating cell free DNA testing: are some test failures informative?

OBJECTIVE: The proportion of circulating cell free DNA derived from the feto-placental unit (fetal fraction or FF) correlates with test success and interpretation reliability. Some fetal disorders are associated with systematically lower FF, sometimes resulting in noninformative results. METHODS: We analyzed results from pregnancies tested in a nested case/control study derived from a cohort of 4664 high-risk pregnancies. Low FF was defined before and after adjusting for maternal weight and gestational age. RESULTS: Compared with euploid pregnancies, the median FF was significantly higher in Down syndrome pregnancies (ratio 1.17) and significantly lower in trisomy 18 and triploid pregnancies (ratios 0.71 and 0.19, respectively). Among 2157 pregnancies tested, 13 (0.6%) had FF <3.0% (all noninformative), including three trisomy 18 and three triploidy fetuses. After adjustment, 16 pregnancies (0.7%) had FF <0.3 multiples of the median (six informative), including one trisomy 18 and three triploidy fetuses. Modeled positive predictive values for low and high-risk populations were 7% and 30%, respectively. CONCLUSION: Among women with noninformative results attributable to low FF, trisomy 18 and/or triploidy risk are sufficiently high to warrant offering additional assessments (e.g. ultrasound). If the testing indication is ultrasound abnormality, amniocentesis and karyotype/microarray should be considered. © 2014 John Wiley & Sons, Ltd.

The reference range and within-person variability of thyroid stimulating hormone during the first and second trimesters of pregnancy.

OBJECTIVE: To further explore first and second trimester reference ranges for thyroid stimulating hormone (TSH) and examine within-person variability of TSH and thyroid peroxidase (TPO) antibody. SETTING: Women coming for routine prenatal care in early pregnancy agreed to participate in a trial of integrated serum screening for Down's syndrome. Two serum samples were obtained from each woman, one each in the first and second trimesters. These samples were also available for TSH and TPO measurements in the present study. METHODS: TSH and TPO antibody measurements were performed in 1126 women with ultrasound-dated pregnancies who provided serum samples in both trimesters. TSH reference ranges were established for the entire cohort and for the antibody-negative subgroup. Within-person variability of TSH measurements between trimesters was examined. RESULTS: Median TSH values are lower in the first trimester than in the second (1.00 versus 1.29 mIU/l), but 98th centile values are higher (5.20 versus 4.18 mIU/l). High correlation exists between individual women's first and second trimester TSH measurements (r=0.75, r2=0.56, p<0.001). Among 23 women with TSH values above the 98th centile in the second trimester, 17 (74%) were over the 95th centile in the first trimester. TPO antibody measurements are also highly correlated between trimesters (r=0.97, r)=0.94). CONCLUSION: Proper interpretation of TSH measurements during pregnancy requires that laboratories establish and monitor appropriate reference ranges. TSH levels show high within-person consistency between trimesters.

Cystic fibrosis: selecting the prenatal screening strategy of choice.

Cystic fibrosis is a serious disorder. Research into the treatment of affected individuals is in progress, but a cure is not expected in the near future. In this review, we demonstrate that prenatal screening for cystic fibrosis meets the requirements for a worthwhile screening programme. We explain the reasons that have led us to conclude that one approach ('couple screening') is the method of choice. The couple-based approach calls for reporting results to the couple as a unit. Only if both parents are found to be carriers is the result designated screen-positive and an amniocentesis or chorionic villus sampling offered. This offers a substantial reduction in the proportion of women with unaffected pregnancies with positive results (the false-positive rate) compared with other methods without reducing the detection of affected pregnancies. It also avoids creating a screen-positive group for which no definitive diagnosis is available. This is a problem with other screening methods. The couple method can achieve a 72% detection rate for a 0.1% false-positive rate. The screening method is simple, non-invasive, reliable, safe and reasonably cost effective. Existing programmes have shown that screening using this method is acceptable to health care professionals and patients. Setting up a national prenatal screening programme for cystic fibrosis is timely and should be implemented using the couple screening method.

Hereditary haemochromatosis and hepatocellular carcinoma in males: a strategy for estimating the potential for primary prevention.

OBJECTIVES: Homozygosity for the C282Y mutation of the HFE gene is the main cause of iron overload in hereditary haemochromatosis. This study calculated the number of hepatocellular carcinoma cases among a cohort of white males that could be attributed to C282Y homozygosity. A better understanding of the extent of potentially preventable mortality arising from this cancer might help with decision making about the feasibility of population screening. METHODS: We combined information from published life tables, age-specific cancer rates and DNA studies of archived liver biopsy specimens to calculate the number of cases of hepatocellular carcinoma that might occur during the lifetime of a cohort of 1,000,000 men, including a subgroup of 5000 C282Y homozygotes. RESULTS: Hepatocellular carcinoma was estimated to occur in 2673 men in the cohort (1:374); 267 of these cases were in the subgroup of 5000 C282Y homozygotes (1:17). If these 267 cases were prevented, the remaining lifetime risk among all males would be 1:416. The relative risk for this cancer in C282Y homozygotes is 23. CONCLUSIONS: There continues to be uncertainty about the efficacy of screening for haemochromatosis. Hepatocellular carcinoma is the most readily quantifiable serious health problem attributable to this source. Further confirmatory DNA (C282Y) studies would be helpful in larger, unbiased sets of archived biopsy specimens, as a way to confirm the present estimate. Any strategy designed to prevent attributable liver cancer is likely to prevent other serious problems from haemochromatasis as well.

Iodine sufficiency and measurements of thyroid function in maternal hypothyroidism.

OBJECTIVES: To test the hypothesis that thyroglobulin (Tg) and free T4 (FT4) concentrations more than 2SD from the control mean are not increased in pregnancy in an iodine replete area in the absence of elevated TSH concentrations. The second hypothesis to be tested was that if such abnormalities in FT4 and Tg in the absence of elevated TSH concentrations were to exist they would not be associated with lowered IQs in the progeny. DESIGN: Cross-sectional study in New England comparing TSH, Tg, antibodies to Tg and FT4 in volunteer nonpregnant women 20-40 years old with those in hypothyroid mothers and matched euthyroid control mothers. The results are contrasted with those from similar studies reported from iodine deficient areas. SUBJECTS: Sera obtained at 17 weeks gestation and stored at -20 degrees C for 8 years were retrieved and analysed from 62 mothers with subclinical hypothyroidism and 124 matched euthyroid mothers. The diagnosis of hypothyroidism was made by finding a TSH concentration > 97.7 percentiile for 25 000 consecutive pregnant women. Sera were also analysed from 53 healthy nonpregnant volunteer women aged 20-40 years. MEASUREMENTS: TSH, Tg and Tg antibodies were measured in the sera of the nonpregnant volunteers, and Tg and Tg antibodies in the sera of the pregnant women who had previously been analysed for TSH and FT4. The incidence of FT4 concentrations below the 2.3 percentile of nonpregnant laboratory controls was compared for the euthyroid and hypothyroid mothers and the laboratory normal controls. RESULTS: Thirty-one per cent of the 62 hypothyroid mothers had FT4 concentrations below the 2.3 percentile compared with only one (0.8%) of the euthyroid mothers. Mean Tg concentrations did not differ between the nonpregnant controls and the euthyroid pregnant women, 14 +/- 10 vs. 16 +/- 10 micro g/l. Tg concentration in the hypothyroid mothers was 44 +/- 61, significantly greater than for either of the euthyroid control groups, P < 0.005. Positive antibodies to Tg were found in 9% and 10% of the control groups and 57% of the hypothyroid mothers, P < 0.0005. When TSH is included as an independent variable in multiple linear and logistic regressions, FT4 and Tg no longer correlate significantly with IQs. CONCLUSIONS: The incidences of FT4 concentrations more than 2SD below the control mean and of Tg > 2SD above the control mean are significantly increased in hypothyroid mothers in iodlne-sufficient New England. However, in the absence of elevated TSH concentrations, the incidences of such abnormalities in FT4 and TG are negligible. Indeed, concentrations for FT4, Tg and Tg antibodies for nonpregnant and pregnant controls in our iodine-replete area do not differ significantly from each other or from previously reported normative concentrations with the methods used. Thus, pregnancy in New England neither increases Tg nor lowers FT4 concentrations.

Assigning risk for Smith-Lemli-Opitz syndrome as part of 2nd trimester screening for Down's syndrome.

OBJECTIVES: To design a reliable model in the context of prenatal screening for assigning the risk in an individual pregnancy of Smith-Lemli-Opitz syndrome (SLOS) and assess its performance. SETTING: A 2nd trimester screening programme for Down's syndrome that measures unconjugated estriol (uE3) along with other serum markers. METHODS: Development of individual risk estimates with a trivariate model incorporating measurements of maternal serum uE3, alpha-fetoprotein (AFP), and human chorionic gonadotropin (hCG) in both SLOS and unaffected pregnancies. RESULTS: Population parameters were computed for the three analytes, as were pairwise correlation coefficients and truncation limits, based on an unbiased collection of 29 affected pregnancies. Published parameters were used for unaffected pregnancies. With a cut off level of risk of 1:50, 62% of SLOS pregnancies can be detected by initially identifying 0.34% of unaffected pregnancies as screen positive. About 1 in 90 screen positive pregnancies will be affected. CONCLUSIONS: It is possible to screen for SLOS as an add on to existing 2nd trimester maternal serum screening, if uE3 is already being measured. A large, prospective trial is necessary to determine whether diagnostic testing can be performed in maternal urine or serum rather than amniotic fluid.

Long QT syndrome in children: the value of the rate corrected QT interval in children who present with fainting.

OBJECTIVE: A strategy was evaluated for identifying a proportion of children with long QT syndrome (LQTS) using the rate corrected QT interval (QTc) to systematically evaluate children who faint. METHODS: QTc measurements and rates of fainting for the present analysis are available from families with KVLQT1, HERG, or SCN5A genotypes. QTc distributions in affected and unaffected children were documented and detection and false positive rates were modelled. RESULTS: The mean QTc (SD) in 117 affected children was 0.484 seconds (0.031), and 0.420 seconds (0.021) among 133 unaffected children. At a cut off of 0.49 seconds, QTc measurement will identify 42.5% of affected and 0.1% of unaffected persons with a history of fainting who are alive at the time of testing. Assumptions include a prevalence of 1:5000 for LQTS, 2% mortality with the first arrhythmia, and a rate of fainting of 50% in affected children and 7% in unaffected children. Given these variables, a QTc cut off of 0.49 seconds detects 42 of 200 affected, along with 70 unaffected children out of a population of 1 million. If QTc > or = 0.49 seconds is found in either parent of children with a QTc of 0.44 through 0.48 seconds, another 21 affected and 25 unaffected children will be identified. CONCLUSION: Systematically performing QTc measurements as part of the evaluation of children who faint might optimally identify about one third of patients with LQTS with few false positives and thereby offer an opportunity to prevent some sudden deaths.

Long QT syndrome in children: the value of rate corrected QT interval and DNA analysis as screening tests in the general population.

OBJECTIVE: To evaluate two hypothetical screening strategies for identifying children with long QT syndrome (LQTS), a cause of sudden death in childhood. METHODS: Families with KVLQT1, HERG, or SCN5A genotypes provided electrocardiographic (ECG) data for this analysis. This is the first time such genotype-phenotype information has been available. Using the LQTS genotype, the distributions of QTc in affected and unaffected children were established and screening performance for various QTc cut off points were modelled. The detection rate for DNA mutation analysis was determined from published experience. RESULTS: The mean QTc (SD) was 0.484 seconds (0.031) in 117 affected children and 0.420 seconds (0.021) in 133 unaffected children. A QTc cut off of 0.50 seconds in a population of 1 million children would identify 61 of the 200 affected children, and 100 unaffected children. Estimates of testing costs for a screening programme in the newborn period would be $327 869/case detected and $2 222 000/death avoided. Although not presently available for routine use, DNA analysis could, theoretically, identify 100 of the 200 children with LQTS within the same population, along with an estimated 100 unaffected children. CONCLUSION: The only available screening test for LQTS is ECG measurement. If DNA technology becomes available for screening, unit costs must be very low to be competitive. There are multiple problems with screening for LQTS: only a minority of children will be detected, cost/death avoided is high, and pilot studies would need to be in place for 5-10 years to document efficacy.

Relation of severity of maternal hypothyroidism to cognitive development of offspring.

BACKGROUND: An association between maternal subclinical hypothyroidism and low intelligence quotient (IQ) in the offspring has recently been shown. OBJECTIVE: To provide evidence for the causality of the association by testing the hypothesis that severity of maternal hypothyroidism correlates inversely with IQ of the offspring. METHODS: IQ scores were compared among 8 year old offspring of 124 control mothers whose thyroid stimulating hormone (TSH) concentrations were < 98th percentile of a cohort of 25,000 mothers at 17 weeks gestation, of 28 untreated hypothyroid women whose TSH was between the 98th and 99.85th percentiles, and of 20 untreated women whose TSH concentration was > or = 99.85th percentile. RESULTS: Mean (SD) IQs for each group of children (in ascending order of maternal TSH concentration) were 107 (13), 102 (15), and 97 (14). The difference between the extremes was significant (p = 0.003). The percentage of children with IQs > 1 SD below the control mean was 15, 21, and 50 respectively (p = 0.003). The odds ratio of having an IQ > 1 SD below the control mean, after controlling for socioeconomic status, was 4.7 (p = 0.006) for the third group compared with the controls. CONCLUSIONS: The inverse correlation between severity of maternal hypothyroidism and IQ of the offspring supports a causal relation and makes the need to screen for and treat pregnant women for hypothyroidism even more compelling.

Clinical validation of a new dimeric inhibin-A assay suitable for second trimester Down's syndrome screening.

OBJECTIVE: To compare the Down's syndrome screening performance of a simplified dimeric inhibin-A assay (Diagnostic Systems Laboratories (DSL)) with an assay whose clinical utility has been established (Serotec). SETTING: A case control set consisting of 51 Down's syndrome and 245 matched unaffected pregnancies collected as part of an earlier multicentre cohort study. METHODS: Sera were assayed for dimeric inhibin-A using the DSL assay and Serotec reference assay. Data analysis included a method comparison of mass values, fit of data to a logarithmic Gausian distribution, and determination of detection and false positive rates. In addition, 234 fresh sera were assayed using the simplified method. RESULTS: The two assays showed a high correlation (r = 0.93) but average concentrations of the DSL assay were 48% higher. However, the differences were basically proportional over the range of values important for screening. The detection rate was essentially equivalent for the DSL assay whether analysed univariately or in combination with other markers (for example, 79% v 75% at a 5% false positive rate for the DSL and Serotec assays for the combination of alpha fetoprotein, unconjugated oestriol, human chorionic gonadotrophin, and dimeric inhibin-A, respectively). The 234 dimeric inhibin-A values measured on fresh sera fitted a logarithm Gaussian distribution for the DSL assay, as indicated by the fit to a probability plot. CONCLUSIONS: The Down's syndrome screening performance of a simplified dimeric inhibin-A immunoassay was equivalent to a more labour intensive established dimeric inhibin-A assay.

Maternal thyroid deficiency and pregnancy complications: implications for population screening.

OBJECTIVE: To examine the relation between certain pregnancy complications and thyroid stimulating hormone (TSH) measurements in a cohort of pregnant women. METHODS: TSH was measured in sera obtained from women during the second trimester as part of routine prenatal care. Information was then collected about vaginal bleeding, premature delivery, low birthweight, abruptio placentae, pregnancy induced hypertension, need for cesarean section, low Apgar scores, and fetal and neonatal death. RESULTS: Among 9403 women with singleton pregnancies, TSH measurements were 6 mU/l or greater in 209 (2.2%). The rate of fetal death was significantly higher in those pregnancies (3.8%) than in the women with TSH less than 6 mU/l (0.9%, odds ratio 4.4, 95% confidence interval 1.9-9.5). Other pregnancy complications did not occur more frequently. CONCLUSION: From the second trimester onward, the major adverse obstetrical outcome associated with raised TSH in the general population is an increased rate of fetal death. If thyroid replacement treatment avoided this problem this would be another reason to consider population screening.

Second-trimester maternal serum analyte levels associated with fetal trisomy 13.

The aim of this study was to determine whether pregnancies affected by fetal trisomy 13 are associated with second-trimester maternal serum analyte levels different from those typical of the unaffected population. Pregnancies with trisomy 13 were identified through cytogenetics laboratories. Those which had second-trimester maternal serum screening analyte measurements were further evaluated. Maternal serum analyte levels for each case and five matched controls were statistically analysed by matched ranked-sum analysis. 28 cases of fetal trisomy 13 were identified. The median AFP, uE3 and hCG levels were 1.35 MoM, 0.71 MoM and 0.90 MoM, respectively. Only uE3 levels were statistically different (p < 0.01) from those for the unaffected population. These data suggest that second-trimester maternal serum AFP, uE3 and hCG levels are not useful in detecting fetal trisomy 13 and protocols already existing for Down syndrome or trisomy 18 screening will not detect the majority of cases of this aneuploidy.

Maternal thyroid deficiency during pregnancy and subsequent neuropsychological development of the child.

BACKGROUND: When thyroid deficiency occurs simultaneously in a pregnant woman and her fetus, the child's neuropsychological development is adversely affected. Whether developmental problems occur when only the mother has hypothyroidism during pregnancy is not known. METHODS: In 1996 and 1997, we measured thyrotropin in stored serum samples collected from 25,216 pregnant women between January 1987 and March 1990. We then located 47 women with serum thyrotropin concentrations at or above the 99.7th percentile of the values for all the pregnant women, 15 women with values between the 98th and 99.6th percentiles, inclusive, in combination with low thyroxine levels, and 124 matched women with normal values. Their seven-to-nine-year-old children, none of whom had hypothyroidism as newborns, underwent 15 tests relating to intelligence, attention, language, reading ability, school performance, and visual-motor performance. RESULTS: The children of the 62 women with high serum thyrotropin concentrations performed slightly less well on all 15 tests. Their full-scale IQ scores on the Wechsler Intelligence Scale for Children, third edition, averaged 4 points lower than those of the children of the 124 matched control women (P= 0.06); 15 percent had scores of 85 or less, as compared with 5 percent of the matched control children. Of the 62 women with thyroid deficiency, 48 were not treated for the condition during the pregnancy under study. The full-scale IQ scores of their children averaged 7 points lower than those of the 124 matched control children (P=0.005); 19 percent had scores of 85 or less. Eleven years after the pregnancy under study, 64 percent of the untreated women and 4 percent of the matched control women had confirmed hypothyroidism. CONCLUSIONS: Undiagnosed hypothyroidism in pregnant women may adversely affect their fetuses; therefore, screening for thyroid deficiency during pregnancy may be warranted.