Screen‐positive rate in cell free DNA screening for trisomy 21

To assess whether the fetal fraction (FF) has an impact on the screen‐positive rate (SPR) in cell‐free DNA (cfDNA) screening for trisomy 21.


| INTRODUCTION
For more than a decade, non-invasive prenatal testing (NIPT), based on the analysis of cell-free DNA (cfDNA) in maternal blood, has been used as a highly sensitive and specific non-invasive method for screening for fetal trisomy 21, 18, 13.The detection rate is about 99% for a false-positive rate of 0.1%.[6] Many maternal organs, especially the hematopoietic system, are the source of circulating cfDNA. 7In pregnancy, trophoblasts also release cfDNA. 8The FF refers to the proportion of cfDNA in the maternal circulation that is derived from the trophoblasts.There are numerous covariates that impact the level of the FF. 9 The most important factors appear to be maternal weight or body mass index, gestational age, the presence or absence of a chromosomal abnormality, and the specific type of the chromosomal abnormality. 9 some studies dealing with screening for trisomy 21 using cfDNA, a correlation was found between the FF and test performance. 10If the FF is below a certain threshold, most cfDNA tests are uninformative.However, even above this threshold, an improvement in the test performance with increasing FF has been noted. 11In a theoretical model, Wright et al. postulated that as the FF increases from 4% to 8%, the detection rage for trisomy 21 increases from 62% to almost 100%.Benn and Cuckle observed a similar effect of the FF on the test performance in screening for microdeletions. 12In a study of more than 52,000 euploid pregnancies and 309 fetuses with a cfDNA result at high risk of microdeletion 22q11.2,we observed a significant increase in the SPR in cases where the FF was low. 13If the FF was below 11%, the SPR was 0.92%, whereas it was 0.13% in cases where the FF was above this level.However, this study was not designed to evaluate the detection and false positive rates.
In the current study, we retrospectively investigated the influence of the FF on the SPR in screening for trisomy 21 using cfDNA in a large, prospectively collected cohort of around 400,000 samples.
The size of the data set forced us to analyze the SPR to evaluate the test performance instead of the actual detection and false-positive rates.Direct evaluation of outcomes was not possible due to the large size of the data set.

| METHODS
This study is based on samples that were analyzed by Cenata or Zentrum für Humangenetik, private laboratories for noninvasive prenatal testing in Tübingen, Germany.[16][17] In brief, this is a targeted cfDNA analysis performed using microarray quantitation of Digital ANalysis of Selected Regions (DANSR) assays of non-polymorphic (chromosomes 13, 18, 21, X and Y) and polymorphic loci (chromosomes 1-12) to estimate chromosome proportion and FF, respectively.The FORTE (FF Optimized Risk of Trisomy Evaluation) algorithm is used to provide patient-specific risk assessments for trisomy.
The measurement of the FF is performed using an SNP-based approach as described by Sparks et al. and Juneau et al. 14,15 In short, the measurement is based on the DANSR assays designed against polymorphic loci on chromosomes 1-12.The assays are unbiased, cover each chromosome uniformly and target SNPs with high minor-allele frequencies in the HapMap dataset to maximize the number of informative loci for fetal-fraction measurement. 17Relative allele frequencies are used to compute FF with the minor percentage of the fraction assumed to be fetal.
Our study cohort comprised of samples from women of at least 16 years of age.Blood was collected in two 10 mL blood collection tubes containing a stabilizer for cfDNA and sent to Cenata or Zentrum für Humangenetik within 7 days after collection at room temperature without prior processing.All women were counseled prior to the cfDNA analysis and after testing according to national regulations.
CfDNA tests with a high-risk result for a chromosomal abnormality other than trisomy 21 were excluded from this study.Twin pregnancies and cases with an uninformative test result were also excluded.
We decided to evaluate the SPR rather than the detection and the false-positive rates.This is due to the fact that our study population included around 400,000 cfDNA examinations.We could not find a reliable way to do a follow-up in these patients to determine the detection and false-positive rates in such a large population.
Data analysis was done as part of a continuing quality assurance program of the two participating laboratories and an anonymized dataset was used.Therefore, approval from the ethical committee of the University Hospital of Tübingen was not required.

| Statistical analysis
Statistical analysis was performed using SPSS 26 (Armonk, New York, USA) and Microsoft Excel (Redmond, Washington, USA).Data are presented as median (25-75th interquartile range) and percentage, whatever was appropriate.
We used uni-and multivariate regression analysis to examine the association between the FF, the gestational age, maternal age, and maternal weight.Based on the regression analysis in the population with a normal cfDNA result, we computed the expected FF for each woman.The FF MoM was calculated as the observed/expected FF.
Univariate regression analysis was also used to assess the relationship between the SPR and the gestational age, maternal age, and maternal weight.
The expected number of screen-positive cases was calculated as follows: we used the number of cases with a normal cfDNA test result and multiplied it with the expected false-positive rate of 0.12%. 3The number of cases with a high-risk result for trisomy 21 was multiplied with the expected detection rate of 98.7%. 3 The sum of both was used as the expected SPR and compared with the observed SPR.
A p-value of <0.05 was used as the significance level. -153 The search of the database identified 401,907 examinations performed between 2018 and 2023.37,026 (9.2%) cases were excluded.
In 24,205 (6.0%) pregnancies, the test was uninformative and in 3071 cases, the test was not performed primarily due to low sample volume.In 2293 (0.6%) cases, the cfDNA test indicated a sex chromosomal abnormality or trisomy 18/13, and 7457 (2.0%) cases were twin gestations.The total number of examinations included in the study was 364,881.
Median maternal and gestational ages were 34.6 (31.4-37.7)years and 12.4 (11.6-13.3)weeks, respectively.Median maternal weight was 66.0 (59.5-75.2) kg.In 362,267 pregnancies, the cfDNA screening result was normal and in 2614 (0.7%) cases, the test indicated high-risk for trisomy 21.Further details of the study population are given in Table 1.
The distributions of FFs in the group with a normal test result and in those with a cfDNA test indicating high-risk for trisomy 21 are shown in Figure 1.
The median FF in the cfDNA tests with a normal and a high-risk result for trisomy 21 result was 9.6% (7.4-12.4) and 10.1% (7.7-13.1),respectively.The FF stratified according to the maternal weight, age and gestational age is shown in Figure 2. Univariate and multiple regression analysis indicated that the FF was dependent on gestational age (OR 0.201, p < 0.001), maternal age (OR 0.022, p < 0.001), and maternal weight (OR −0.091, p < 0.001) (Table 2).The SPR for the entire population was 0.72%.Figure 3  98.1% of the study population.In cases with >20% FF, the positive association between SPR and increasing FF was much stronger (OR 1.099, p < 0.001, Nagelkerkes r2 = 0.042).This association was still evident when the MoM distribution was used, but was less pronounced compared to the absolute FF.The increase in the SPR is related to the excess of screen-positive cases in the late second and third trimesters.
According to the maternal and gestational age distribution of our study population, we estimated that our study population should include 1993.0 fetuses with trisomy 21.
In order to compare the observed SPR with the expected rate, we estimated the expected number of screen-positive cases in screening for trisomy 21 as follows: The cfDNA tests with a normal test result were multiplied with a false-positive rate of 0.12% (n = 434.7).We applied a detection rate of 98.7% to the trisomy 21 test results, corresponding to n = 1967.1 cases.In summary, the total expected number of screen-positive cases, was 434.7 þ 1967.1 = 2401.8(95% prediction interval 2305.7-2497.9).
As there were 2614 screen-positive cases, the observed SPR was 8.8% higher than expected.Table 3 and Figure 6 demonstrates the estimated and observed screen-positive rates according to the FF.
Up to a FF of 8%, we observed 712 screen-positive cases and expected 791.3 screen-positive pregnancies.Thus, the SPR was 0.62% or 11.1% lower than expected.Above this threshold until a FF of 20%, there were 1902 observed and 1610.5 expected cases.The SPR was 0.74% and 18.1% higher than expected.In the group of pregnancies with a FF above 20%, the SPR was 1.55% or 147.2% higher than expected (expected positive cases: n = 44.5, observed n = 110) (Figure 6).

| DISCUSSION
In this study, we investigated the impact of FF on the SPR in screening for trisomy 21 using cfDNA.We used the SPR as a surrogate marker for the test performance as it was impossible to obtain outcomes for such a large study population.Our study shows that the FF did not have a significant impact on the SPR in screening for trisomy 21.The SPR was constant starting with FF of about 0.7% up to FF of 20%, which includes about 98% of our study population.
We also compared the observed with the expected SPR and found that the observed rate was 8.8% higher than expected but, from a clinical point of view, within an acceptable range.Up to a FF of 7%, the SPR was at the lower limit of the expected range (Figure 6).
In the group of pregnancies with a higher FF, especially with more than 20%, the SPR was higher than estimated.We found a similar increase in the SPR with increasing gestational age, especially after 20 weeks.Therefore, we assume that in pregnancies where fetal defects or soft markers were not detected until the second or third trimester, many women opted for a cfDNA test instead of an invasive test, as the result did not influence their decision to continue with the pregnancy.
The fact that many women opt against the termination of pregnancy in case of a chromosomal abnormality was shown by others before.In a previous study from our group, we observed an overall termination rate of 76% after the diagnosis of trisomy 21.However, the termination rate highly depended on gestational age.Less than 20% chose termination if a chromosomal abnormality was identified in the late second and third trimester. 18 test that was used in our study increases the odds ratio for the differentiation between affected and unaffected pregnancies at low FF. 15,16However, in screening for microdeletion 22q11.2 with the same cfDNA test, we found that the FF had a significant impact on SPR, which was clinically relevant. 13In the current study, the FF did not have a significant impact on the SPR in screening for trisomy 21.
We acknowledge that our study has some limitations.The main one is that due to the high number of patients, we had to evaluate the SPR instead of the detection and false-positive rate.The expected SPR can only be estimated, and we acknowledge that this approach has some weaknesses.Instead of using the cohort of fetuses that were known to be euploid and those with trisomy 21, we resorted to using the screen-positive and screen-negative groups and multiplied these numbers with the anticipated detection and false-positive rate.
For the detection and false-positive rate, we used the recent metaanalysis from Demko et al 3 that did not perfectly match the performance data of the Harmony ® test.Nevertheless, from the fact that the observed SPR is so close to the expected rate, we can conclude that screening for trisomy 21 works reasonably well for any FF above 4%.
In conclusion, we have demonstrated, that the SPR in cfDNA screening for trisomy 21 was relatively constant up to a FF of about 20% and within the expected range.
Open Access funding enabled and organized by Projekt DEAL.

F
shows the SPR stratified according to the maternal age, maternal weight, and gestational age.Figures 4 and 5 demonstrate the SPR stratified according to the absolute FF and the FF MoM values.The SPR was independent of maternal weight but dependent on maternal age (maternal weight: OR 1.000, p = 0.894, Nagelkerkes r2 = 0.000; maternal age: OR 1.186, p < 0.001, Nagelkerkes r2 = 0.048).There was also a positive but weak association between the FF and SPR in the cases where the FF was ≤20.0%(OR p = 1.021, p < 0.001, Nagelkerkes r2 = 0.001).This group included 357,800 pregnancies or F I G U R E 2 Fetal fraction stratified according to the maternal age (left), maternal weight (middle), and gestational age (right).T A B L E 2 Uni-and multivariate regression analysis between the fetal fraction and the maternal weight and age as well as the gestational age.I G U R E 3 Screen-positive rate stratified according to the maternal age (left), maternal weight (middle), and gestational age (right).The lines represent the 95% confidence intervals.[Colour figure can be viewed at wileyonlinelibrary.com]LÜTHGENS ET AL.F I G U R E 4 Screen-positive rate stratified by fetal fraction.The lines represent the 95% confidence intervals.F I G U R E 5 Screen-positive rate stratified by fetal fraction MoM.The lines represent the 95% confidence intervals.[Colour figure can be viewed at wileyonlinelibrary.com] 1540 -LÜTHGENS ET AL.

F I G U R E 6
Observed (red circles) and expected (black lines) screen-positive rates according to the fetal fraction (FF).The upper and lower black lines represent the 95% prediction interval for the expected FF.[Colour figure can be viewed at wileyonlinelibrary.com]

Median fetal fraction (%) Normal cfDNA test results n Expected trisomy 21 cases n Expected false- pos cases (FPR 0.12%) n Expected detected cases (DR 98.7%) n Total n 95% prediction interval n Observed screen-positive cases n of
Observed and expected screen-positive cases for trisomy 21 stratified by fetal fraction.
19e women opted against termination of pregnancy.19It is reasonable to assume that a woman who has decided to continue with the pregnancy, even if a fetal defect or soft markers indicate trisomy 21, will opt for NIPT instead of invasive testing.