Noninvasive Prenatal Testing and Detection of Maternal Cancer

Abstract

Noninvasive prenatal testing (NIPT) to detect fetal aneuploidy (an abnormal number of chromosomes) usingmassive parallel sequencing of cell-free DNA (cfDNA) from maternal blood is gaining rapid acceptance in obstetrics, given its high sensitivity and specificity for the detection of trisomies 21, 18, and 13. Placental DNA, a surrogate of fetal DNA, enters the maternal bloodstream via apoptosis of trophoblasts in the intervillous space of the placenta,where itmixes with cfDNA of maternal origin. The fetal fraction in this mixture usually varies between 3% and 15% in the late first and second trimesters.NIPT represents a formofadvancedscreening, and positive results consistent with aneuploidy require confirmation by a diagnostic test usingmaterial obtained via amniocentesis or chorionic villus sampling (CVS).1,2 The sensitivity of NIPT for the detection of trisomies 21, 18, and 13 ranges between 91% and 99.2%; the false-positive rates range between 0.09% and 0.13%.3 False-positive results havebeen attributed to confinedplacentalmosaicism,maternal chromosomalmosaicism,maternalDNAcopynumbervariants, andmaternal organ transplant fromamale donor.4 Two case reports (in 2013 and 2015) suggested that the discordant results (between those of NIPT and amniocentesis) could be due to undiagnosed maternal cancer.5,6 The first case described a 37-year-old pregnant woman with an uncomplicated pregnancy who had NIPT performed at 13 weeks of gestation. The findings were suggestive of fetal aneuploidy for chromosomes 13 and 18 in 2 different maternal blood samples. Yet the fetal karyotype was 46,XY without evidence of submicroscopic genomic imbalances with wholegenome oligonucleotide microarray. After delivering a healthy infant, the patient reported pelvic pain found to be due to a pathologic fracture of the pubis; a biopsy showed metastatic neuroendocrine carcinoma of unknown origin. A separate lesion on the anterior vaginal wall proved to be a high-grade neuroendocrine small cell carcinoma. Tumor cells showed aneuploidy for chromosomes 13 and 18, consistent with the findings from the analysis of maternal blood assessed with NIPT.5 The second patient was a 27-year-old asymptomatic woman at 12 weeks of gestation whose NIPT results had abnormal genomic findings for chromosomes 8, 9, and 14 that were discordantwith a normal fetal karyotype from amniotic fluid cells. Moreover, maternal lymphocytes did not have evidence of aneuploidy ormosaicism. Given the genomic imbalances, amalignancywas suspected, and thepatientunderwent a whole-body diffusion magnetic resonance imaging (MRI) scan, which showed an anterior mediastinal mass and enlarged lymph nodes in the cervical and retroclavicular regions. The diagnosis of nodular sclerosis Hodgkin lymphoma wasmade,basedonabiopsyof themediastinalmass.Hodgkin/ Reed-Sternbergcells showedfindingsconsistentwith thoseobservedwith the analysis of cfDNA inmaternal blood.Basedon these observations, the authors investigated whether cfDNA couldbe informative innewlydiagnosedcasesofnodular sclerosis Hodgkin lymphoma in nonpregnant women (n = 4) and in nonpregnant patients who had a first relapse after therapy (n = 5). The results in 8 of the 9 cases showed genomic imbalances, and such findingswere validated by fluorescent in situ hybridization on Hodgkin/Reed-Sternberg cells in tissue specimens.6 Commonfeaturesof the2caseswere that thepatientswere youngasymptomaticpregnantwomen,hadabnormalNIPT results discrepant with the fetal karyotype (“false-positive NIPT”),were subsequently diagnosed tohave cancer, and the malignant cells showedevidenceofmultiple aneuploidies that mirrored the findings reported by maternal cfDNA analyzed for NIPT. A recent study reported that amongmore than4000pregnant women undergoing NIPT, 3 patients were found to have aberrant genome representation profiles that were reminiscent of cancer-related copy number variation. These women underwent whole-body diffusion-weighted MRI and were found to have cancer—the first had a high-grade serous ovarian carcinoma (stage IV with multiple metastases); the second, a follicular lymphoma; and the third, Hodgkin lymphoma. Analyses of the genome representation profile of the tumors showed that the copy-number variations detected in maternal plasma were present in the tumors.7 The report by Bianchi et al8 in this issue of JAMA is the first large-scale study that strengthens the premise that a false-positive NIPT result could be due to the presence of maternal malignancy based on the results of a case series of pregnant women who had NIPT and were subsequently diagnosed with cancer. This work leverages the information of a laboratory that had processed 125 426 samples from NIPT over 2.5 years with a test to detect the presence of aneuploidy for chromosomes 21, 18, and 13 (testing for sex chromosome aneuploidy was optional). Of these samples, 3% (3757) had NIPT results consistent with aneuploidy. The practice of the laboratory was to discuss all abnormal NIPT results with the referring clinician and recommend a diagnostic test, and clinicians were asked to inform the laboratory if the results Related article page 162 Opinion