Positive predictive values and outcomes for uninformative cell-free DNA tests: An Italian multicentric Cytogenetic and cytogenomic Audit of diagnOstic testing (ICARO study).

OBJECTIVES: To establish the positive predictive values (PPV) of cfDNA testing based on data from a nationwide survey of independent clinical cytogenetics laboratories. METHODS: Prenatal diagnostic test results obtained by Italian laboratories between 2013 and March 2020 were compiled for women with positive non-invasive prenatal tests (NIPT), without an NIPT result, and cases where there was sex discordancy between the NIPT and ultrasound. PPV and other summary data were reviewed. RESULTS: Diagnostic test results were collected for 1327 women with a positive NIPT. The highest PPVs were for Trisomy (T) 21 (624/671, 93%) and XYY (26/27, 96.3%), while rare autosomal trisomies (9/47, 19.1%) and recurrent microdeletions (8/55, 14.5%) had the lowest PPVs. PPVs for T21, T18, and T13 were significantly higher when diagnostic confirmation was carried out on chorionic villi (97.5%) compared to amniotic fluid (89.5%) (p < 0.001). In 19/139 (13.9%), of no result cases, a cytogenetic abnormality was detected. Follow-up genetic testing provided explanations for 3/6 cases with a fetal sex discordancy between NIPT and ultrasound. CONCLUSIONS: NIPT PPVs differ across the conditions screened and the tissues studied in diagnostic testing. This variability, issues associated with fetal sex discordancy, and no results, illustrate the importance of pre- and post-test counselling.

Validation of a method for noninvasive prenatal testing for fetal aneuploidies risk and considerations for its introduction in the Public Health System.

OBJECTIVE: The aim of this study was to validate noninvasive prenatal testing (NIPT) for fetal aneuploidies by whole-genome massively parallel sequencing (MPS). METHODS: MPS was performed on cell-free DNA (cfDNA) isolated from maternal plasma in two groups: a first set of 186 euploid samples and a second set of 195 samples enriched of aneuploid cases (n = 69); digital PCR for fetal fraction (FF) assessment was performed on 178/381 samples. Cases with <10 × 106 reads (n = 54) were excluded for downstream data analysis. Follow-up data (invasive testing results or neonatal information) were available for all samples. Performances in terms of specificity/sensitivity and Z-score distributions were evaluated. RESULTS: All positive samples for trisomy 21 (T21) (n = 43), trisomy 18 (T18) (n = 6) and trisomy 13 (T13) (n = 7) were correctly identified (sensitivity: 99.9%); 5 false positive results were reported: 3 for T21 (specificity = 98.9%) and 2 for T13 (specificity = 99.4%). Besides FF, total cfDNA concentration seems another important parameter for MPS, since it influences the number of reads. CONCLUSIONS: The overall test accuracy allowed us introducing NIPT for T21, T18 and T13 as a clinical service for pregnant women after 10 + 4 weeks of gestation. Sex chromosome aneuploidy assessment needs further validation due to the limited number of aneuploid cases in this study.

The Val142Ile transthyretin cardiac amyloidosis: not only an Afro-American pathogenic variant? A single-centre Italian experience.

Transthyterin amyloidosis is a life-threatening disorder caused by the deposition of hepatocyte-derived transthyretin (TTR) amyloid in various tissues and organs. The most common worldwide pathogenic variant with almost exclusive cardiac involvement is Val142Ile with an allele frequency of 3.5% in U.S. African-American population, but supposed extremely rare, with only sporadic cases in Caucasian patients. Unexpectedly, in our amyloidosis referral centre, we identified five patients (15.1% of all TTRm diagnosed patients, three families, two singleton) with Val142Ile variant belonging to unrelated families of Caucasian origin. Molecular study was performed in a total of 10 individuals of which three were Italian families (three affected individuals and five unaffected individuals) and two were singleton (one Italian patient and one patient from Argentine with Spanish ancestry). Sequence analysis of TTR gene revealed the presence of the heterozygous Val142Ile in the five affected patients and in five asymptomatic individuals. All probands underwent, at diagnosis, a complete clinical, echocardiographic and biohumoral evaluation. To the best of our knowledge, we describe the larger report of Caucasian patients with Val142Ile cardiomyopathy. All patients at diagnosis showed symptoms of heart failure with increased thickness of left ventricular walls and systo-diastolic left ventricular dysfunction. They also showed increased plasma values of NT-proBNP and troponin I. Our data confirm that Caucasian patients with the Val142Ile pathogenic variant have phenotypic manifestations similar to that of African-American one. Moreover, our data clearly show that Val142Ile pathogenic variant is not only an African-American mutation but could be also an underestimated Caucasian variant.

A Systematic Assessment of Accuracy in Detecting Somatic Mosaic Variants by Deep Amplicon Sequencing: Application to NF2 Gene.

The accurate detection of low-allelic variants is still challenging, particularly for the identification of somatic mosaicism, where matched control sample is not available. High throughput sequencing, by the simultaneous and independent analysis of thousands of different DNA fragments, might overcome many of the limits of traditional methods, greatly increasing the sensitivity. However, it is necessary to take into account the high number of false positives that may arise due to the lack of matched control samples. Here, we applied deep amplicon sequencing to the analysis of samples with known genotype and variant allele fraction (VAF) followed by a tailored statistical analysis. This method allowed to define a minimum value of VAF for detecting mosaic variants with high accuracy. Then, we exploited the estimated VAF to select candidate alterations in NF2 gene in 34 samples with unknown genotype (30 blood and 4 tumor DNAs), demonstrating the suitability of our method. The strategy we propose optimizes the use of deep amplicon sequencing for the identification of low abundance variants. Moreover, our method can be applied to different high throughput sequencing approaches to estimate the background noise and define the accuracy of the experimental design.

Phase II trial of customized first line chemotherapy according to ERCC1 and RRM1 SNPs in patients with advanced non-small-cell lung cancer.

OBJECTIVES: Customized chemotherapy has several advantages: patients are more likely to be treated with the most effective agents and can be spared the toxicity of ineffective drugs. Based on the literature, excision repair cross complementation group 1 (ERCC1) and ribonucleotide reductase M1 (RRM1) genes represent predictive biomarkers of response to platinum compound and gemcitabine, in NSCLC. MATERIALS AND METHODS: We had planned a phase II trial (Simon design) to evaluate combination chemotherapy according to single nucleotide polymorphisms (SNPs) of ERCC1 (118T/C and 8092C/A) and RRM1 (-37C/A and -524T/C) in naïve patients affected by advanced NSCLC. ERCC1 and RRM1 SNPs assessment was performed in peripheral blood lymphocytes (PBLs). Combination chemotherapy was selected based on ERCC1 and RRM1 SNPs: we assume that patients with one or two C alleles at position 118 and with one or two A alleles at position 8092 in ERCC1 gene would correspond to Cisplatin non-responder and than with two A alleles at -37 and two C alleles at -524 in RRM1 gene to gemcitabine non-responder. Four schedules were provided: cisplatin+gemcitabine, cisplatin+docetaxel, gemcitabine+docetaxel; docetaxel+vinorelbine. Primary endpoint was overall response (ORR) in the intention-to-treat population. RESULTS: 42 patients were enrolled from January 2010 to November 2011; 40 patients received at least 1 cycle of chemotherapy; median age was 66 years (range: 47-72); 36(90%) had stage IV, 4(10%) IIIB; 23(58%) had adenocarcinoma, 14(35%) squamous carcinoma. Twenty-five (62%) patients received treatment A, 3(8%) treatment B, 11(28%) treatment C, 1(23%) treatment D. ORR was 55%, analysis in squamous patients subgroups showed 71.4% ORR. The median follow-up was 19.7 months, PFS was 23 weeks (95% CI = 15-26) and OS was 40.4 weeks (95% CI = 32-55). Treatment was well tolerated. CONCLUSION: We observed an increase of ORR in NSCLC patients when they were treated with chemotherapy according to ERCC1 and RRM1 SNPs status.