Ishwar Chander Verma Serves as the HOD in the Department of Centre of Medical Genetics, Sir Ganga Ram Hospital, New Delhi. He has an experience of 51 years in the field of genetics and added valuable contribution in the field of pediatric genetics and community health in India. He handled approximately 2000 cases with expertise in Fetal Medicine and Reproductive Genetics. Current areas of interest and/or active clinical practices: Management of fetal medicine and reproductive genetics; Dysmorphology; Genetic counseling and prenatal diagnosis using molecular, cytogenetic and biochemical techniques. Official certifications/recognition: Holds Genetics training in UK, USA and Switzerland; Member of Fellow of the Royal College of Physicians (FRCP), London, the American Academy of Pediatrics (FAAP) and the National Academy of Medical Sciences (FAMS), New Delhi; Awarded as Ranbaxy Science Award, the Indian Council of Medical Research (ICMR) award, the National Academy of Medical Sciences (NAMS) award and Dr B C Roy (Medical Council of India) National Awards; Mentioned in the Limca Book of Records 2003 as a pioneer in genetics in India.
Materials and Methods: The PanoramaTM NIPT was performed
in 516 pregnancies, which had tested intermediateto-
high risk on conventional first and second trimester
screening. Results were confirmed either by invasive
diagnostic testing or by clinical evaluation after birth.
Results Of 511 samples analyzed, results were obtained in
499 (97.7%). Of these, 480 (98.2%) were low risk and 19
were high risk. A sensitivity of 100% was obtained for
detection of trisomies 21, 18, 13 and sex chromosomal
abnormalities. The specificity ranged from 99.3 to 100%
for abnormalities tested. Taken together, the positive predictive
value for trisomies 21, 18, 13 and monosomy X was
85.7%. The average fetal fraction was 8.2%, which is lower
than the average observed elsewhere.
Conclusion: This is the first report of detailed experience
with NIPT in India and demonstrates comparable performance
in all aspects of testing to the results elsewhere.
Keywords: Prenatal screening India NIPT SNP
Trisomy 21 Trisomy 18 Trisomy 13
Chromosomal aneuploidies
The incidence of chromosomal disorders in India is 1:166
live births [1]. Given the large population, around 35,000
fetuses with Down syndrome alone are conceived every
year [1]. Therefore, screening and diagnosis for chromosomal
disorders is important in India, as in other countries.
The current prenatal screening for chromosomal abnormalities
consists of analyzing blood hormone levels and
ultrasonography. However, these procedures are limited by low sensitivity and high false positive rate of 2–7% [2, 3].
Also, conventional screening misses over 10% of affected
fetuses [4]. Invasive diagnostic methods, such as amniocentesis
or chorionic villus sampling (CVS), are highly
sensitive but cannot be offered to all pregnant women as
they carry a small but significant risk of miscarriage [5, 6].
NIPT, a recently developed advanced technology provides
a significant improvement over conventional testing, with
detection rate of over 99% and a false positive rate of less
than 0.1% by investigation of cell-free fetal (placental) DNA
from maternal blood [7–11]. Furthermore, a significant
reduction, 50–70%, of invasive procedures has been observed
in setups where NIPT has been implemented [12, 13].
Commercialized NIPT technologies follow either of the
two approaches: a counting-based method using massively
parallel sequencing (MPSS) or the single nucleotide polymorphism-
based approach (SNP), used in the PanoramaTM
NIPT developed by Natera Inc. (San Carlos, USA). The
technology and the performance of the SNP-based method
has been described elsewhere and validated both in high and
low risk pregnant women [14–20]. The advantage of this
method is that it does not require a reference chromosome, is
able to detect vanishing twin, triploidy, maternal mosaicism
and is highly accurate [16]. The limitation has been, rarely,
the inability to make a call when there is a high genetic
homology between parents (consanguinity) [21]. However,
improvement of the algorithm has been implemented, such as
quantitative multiple model (QMM) that resolves samples
with genetic homology [personal communication] and
reduction in no-call threshold for sample calling to 2.8% fetal
fraction [22]. Several professional societies worldwide have
issued guidelines periodically based on available data on
appropriate usage of NIPT [13, 23–27].
The Indian population is socioculturally, ethnically and
biologically diverse [28] as reflected in the studies of
mutations detection that reveal many novel ones. Additionally,
the Indian population has a relatively high rate of
consanguinity (20–39%) among certain communities
[29, 30]. It is not known how these biological factors would
influence the SNP-based NIPT test. The collection and handling of samples in a tropical country might also affect
the test performance. To study the feasibility of performing
NIPT in an Indian population given the above considerations,
we conducted an Indian study. The main objective
was to evaluate the performance of NIPT for trisomies 21,
18, 13, sex chromosome abnormalities and triploidy in a
cohort with intermediate-to-high risk on conventional
screening.
Ten leading institutes collaborated in this study, after review and approval from their respective institutional review boards. The flowchart of the study protocol is shown in Fig. 1. Pregnant women were enrolled based on the specified criteria. Singleton pregnancies, between 11 and 18 weeks gestation (CRL 45–84 mm), intermediate-tohigh risk (risk[1:1000 for trisomy 21 and[1:400 for trisomy 18, Nuchal Translucency (NT) measure\95th centile) on biochemical, combined, triple or quadruple screening were included. Score of[1:250 was defined as high risk and between 1:250 and 1:1000 as intermediate risk. While pregnancies with egg/sperm donor, surrogacy, twin or multiple gestation, with known parental chromosomal abnormalities (including known balanced translocations), where invasive testing was planned, NT[95th percentile or nuchal fold thickness[6 mm, C1 malfor-mation in fetus, bone marrow transplant or malignancypatients were excluded.
Twenty milliliters of peripheral blood was collected inStreckTMtubes after genetic counseling and informedconsent. A paternal buccal swab was obtained whenavailable. Samples were transported to Medgenome Lab-oratory in Bangalore. Laboratory testing was performedusing validated protocol from Natera Inc., using SNPs onfive chromosomes and analyzed by the cloud-based pro-prietary NATUS algorithm [10,14,15,22], the enhancedversion of which became available during the course of thestudy. Fetal fraction was reported in each case. In accor-dance with the legal requirements in India, gender was notrevealed to any study personnel [31]. Pertinent details werecollected from the pregnant women. For women whounderwent invasive test, fluorescence in situ hybridizationand karyotyping was done on fetal material. Redraws wererequested for these reasons: lysis, low sample volume, fetalfraction below threshold (2.8%), failed quality matrices,and algorithm-based outcomes such as low confidenceresults. Follow-up information was obtained from theparticipants till the delivery. Information on false negativeoutcomes was keenly sought.
Sensitivity and specificity for each disorder were cal-culated separately using only confirmed cases afterdiagnostic tests or clinical evaluation after birth. Sam-ples that received a no-callwere not included. Descrip-tive data analysis was performed. Where applicable, thettest was used for statistical analysis, andp\0.05 wasaccepted as significant.
The demographic data did not differ significantly amongthe various groups, except; in the no-call group, thematernal weight and BMI were higher (though statisti-cally not significant), while the fetal fraction was signif-icantly lower as compared with the other two groups(p\0.05) (Table1). The overall results are depicted inFig.2, while details of each high-risk result are shown inTable2. Fourteen (73.7%) of 19 high-risk NIPT calls and323 (67.3%) of 480 of low-risk NIPT calls had a priorhigh risk on conventional screening, while 5 of 19(26.3%) high risk and 157 of 480 (32.7%) low risk werein the intermediate-risk range (p[0.05). No significantassociation was found between a high- or low-risk callwith advanced maternal age (p[0.05). Table3lists thetest performance for each chromosomal abnormality,based on the cytogenetic results; combined positive pre-dictive value (PPV) for all chromosomal abnormalitieswas 81.25%
Follow-up information was received from 477 of 511(93.3%) cases. The uptake for invasive testing was 16(84.2%) of 19 women (Table2). Two women terminatedthe pregnancy without confirmatory testing (2/19, 10.5%),and one woman (1/19, 5.3%) refused invasive test. Thesethree unconfirmed cases of trisomy 21 were suggestive ofabnormality as intrauterine death was reported in one caseand ultrasound abnormalities in the other two. Follow-upinformation was available in 452 (94.2%) of 480 low-riskcases, eleven of which were confirmed normal after inva-sive test (performed for other reasons) and the rest wereclinically normal at birth. Follow-up information wasreceived in 9 (81.8%) of 11 no-call cases (Table4). Nofollow-up was obtained in 28 (5.8%) of 480 cases, whichwere excluded from statistical analysis where applicable.No false negatives were reported till the time ofpublication.
Of 511 cases, 30 (5.9%) redraws were requested due to acombination of pre-analytic and post-analytic factors. Thecauses of redraw were: low sample volume/moderate tosevere lysis (7), low confidence result (10), fetal fractionbelow threshold (5), failed libraries (4), failed qualitycontrol (2), ambient genotype contamination and an unin-formative DNA pattern one each. Pre-analytical factorsaccounted for 23% of the redraws, while the laboratoryprocessing and algorithm-based redraws were 76.7%.Redraws were received in 22 (73.3%) of 30 samples, and adefinitive result was obtained in 19 (86.3%) of 22 redraws.
Most samples (97.2%) werereceived within 48 h. Most(88.8%) samples were reported within 12 days (Fig.3). Theaverage turnaround time for reporting was 7 calendar days.
The average FF in the no-call cases (3.9%) was signifi-cantly lower than in samples that received a call (8.7%)(Welch two-sample t test,p\0.001) (Table1). The lowestFF for which a call was made was 3%. Five (1%) of 511cases were found to have a FF below threshold (2.8%).Redraws were received in 3 of 5 cases with low FF;however, a call on redraw could be made only in one(Table4). The FF was directly proportional to gestationalage and inversely proportional to maternal weight (Fig.4).
The overall no-call rate was 2.3% (12/511). This wasdefined as no result on the original sample (when no redrawis received) or on a redraw. Redraws were received in only3 of the 11 samples, which were no-calls again (Table4).
Information was available in 453 (88.6%) of 511 samplesconfirming that the sample cohort consisted of pregnantwomen of Indian origin.
Of 511 samples, consanguinity information was availablein 479 (93.7%). Of these 17 (3.5%) were consanguineous,and all 17 samples received a call either the regular algo-rithm or QMM. Only 3 (17.6%) of 17 pregnant women whoreported consanguinity yielded an uninformative DNApattern on initial analysis. All three samples received adefinite call by QMM algorithm. No-call due touninformative DNA pattern was observed in 7 (1.3%) of511 (all chromosome no-call—5 cases, and no-call onchromosome X—2 cases) samples. Of these, only one(partial no-call) had history of consanguinity and interest-ingly also screened high risk of trisomy 21. Of the sampleswith total uninformative DNA pattern, 80% (4/5) could beresolved on QMM algorithm.
Obesity (defined as BMI[30) was observed in 39 (13.9%)of 281 pregnant women. Calls were available in 38 (97.4%)cases on the first draw, of which one was high risk oftrisomy 21 (confirmed on invasive testing). Only 2.7% (1/37) did not receive a call on the first run, and a redraw wasrequested but was not received.
Overall, 97.9% samples received a result within 12 days.The good performance in this study is consistent with theoverall performance of NIPT [8,19,20,32]. The overallpositive predictive value of trisomy 21, trisomy 18, trisomy13 and monosomy X was 85.7%, which was comparable toprevious publication—82.9% for the four aneuploidies[23]. However, the PPV for trisomy 21 was 80%, whichwas slightly lower than 90.9% in the clinical validationstudy [20]. If the three unconfirmed high-risk trisomy 21cases were true positives, the PPV would increase to84.6%. Nevertheless, this PPV is much higher than theconventional screening methods [8,33]. The PPV for sexchromosome abnormalities including monosomy X was75%, which is higher than that published previously—48.4% [34]. It is important to note here that PPV is notintrinsic to the test but depends on the prevalence of thecondition in the tested population [35]. An overwhelmingnumber of women (96.6%), intermediate-to-high risk onconventional screening, were low risk on NIPT. Therefore,potentially, these women avoided invasive procedures foraneuploidy confirmation [13]. The high negative predictivevalue is a significant benefit in terms of reassurance topregnant women.
The average fetal fraction in Indian pregnant womenwas 8.3%, which is lower than the average fetal fractionrange published in the Western population (10.69%)[11,19]. These results are, however, consistent with lowerfetal fractions in women of South Asian origin between 11and 14 weeks [36]. This study did not observe negativeimpact of fetal fractions on NIPT calls in obese women,probably because the degree of obesity observed amongIndian women is lower than in the women in other studiedpopulation, and mean BMI observed in the no-call cohort was only 27.3 [37]. The no-call rate observed was identicalto the no-call rate in the validation study [20]. The no-callrates in other studies ranged between 2.9 and 8.1%[8,19,35]. Previous publications and guidelines haveindicated that no-calls are likely to be aneuploid [11,19].From our no-call cases, a suspected T18 could not beconfirmed. In practice, therefore, the decision to resampleor proceed to invasive testing, given a no-call result, shouldbe made based on reason for the no-call, the original fetalfraction, the probability of receiving a result on a redraw,established guidelines and the applicable regulations ofprenatal testing.
Genetic counseling is mandatory, as recommended byseveral professional societies [27]. In the present study, afetus with confirmed 47, XXX, was continued to term aftercounseling, and the baby is doing well postnatally. All other confirmed high-risk pregnancies were terminated.Unfortunately, two cases high risk on NIPT were termi-nated without confirmation. This undesirable outcomeobserved in other studies too remains a challenge [20].
Although consanguinity is cited as one of the reasons forinability to make a call in the SNP-based NIPT [21], it didnot impact the current study. Limitations of the study werethe small sample size, a biased cohort, as samples werepreselected and a lack of complete follow-up (follow-upinformation was received in 93.3% subjects) and theinability to determine the cause of false positives.
Indian Council of Medical Research recommendsgenetic screening services to all pregnant women in theNational and Family Welfare Program [38]. NIPT, in view of its strong performance in the Indian scenario, safety,accuracy and its easy extension to peripheral areas wouldbe a valuable addition to the prenatal screening program,once the cost comes down. The experience gained in thisstudy has enabled us to consider the implications andsuggest modifications to international guidelines beforeapplying these to India [39]. Currently, NIPT may be usedin cases with high-risk results on conventional screening toavoid unnecessary invasive tests. Further reduction in costand greater awareness would provide the benefits of thisremarkable technology more widely.
Conflict of interest: Authors Dr. Ramprasad VL, Dr. Priya Kadam,Dr. Venkataswamy E, Shruti Lingaiah, Riyaz Akhtar, FrancisKidangan, Chandran R., Kiran C., and Ravi Kumar G. R. were/areemployed with Medgenome Laboratories Private Limited during thecourse of the project.
Ethical approval: The study was reviewed and approved by ethicalreview board of the institutions that participated in the study. Allprocedures performed in studies involving human participants were inaccordance with the ethical standards of the institutional and/ornational research committee and with the 1964 Helsinki declarationand its later amendments or comparable ethical standards.
Informed Consent: Informed consent was obtained from all indi-vidual participants included in the study.
