Portal de Periódicos da CAPES

Clinical implementation of cell‐free DNA ‐based aneuploidy screening: perspectives from a national audit

2014; Wiley; Volume: 45; Issue: 1 Linguagem: Inglês

10.1002/uog.14699

1469-0705

Lisa Hui, Mark Teoh, Fabrício da Silva Costa, Philippa Ramsay, Ricardo Palma‐Dias, Zara Richmond, Sofie Piessens, Susan P. Walker,

Parvovirus B19 Infection Studies

In late 2011, a prenatal screening test for fetal chromosomal abnormalities using cell-free DNA in maternal plasma was introduced commercially in the USA. This next-generation sequencing-based method, commonly referred to as non-invasive prenatal testing (NIPT), represented the most accurate form of screening for trisomy 21 to date1. While the NIPT market expanded rapidly in the USA, its clinical implementation in other developed countries varied considerably due to local factors, such as existing care models, insurance coverage and legal restrictions. In late 2012, NIPT became available clinically in Australia, through overseas laboratories, 1 year after it did in the USA. By the end of 2013, there were five providers in the Australian market offering NIPT on a self-funded basis. Australian subspecialists in maternal–fetal medicine and obstetric ultrasound quickly became the major sources of referral for NIPT due to their well-established role in first-trimester screening and prenatal diagnostic procedures. Despite being 'early adopters' of technology, Australian sonologists had concerns about NIPT that were common to many countries. Among these were definition of the appropriate indications for use, and the uncertain test-failure rates and turnaround times associated with offshore laboratory processing outside trial conditions. The lack of government regulation and lack of data collection were also key concerns2. At the time, international reports on clinical implementation were either single-center experiences3, 4, or multicenter industry-sponsored studies of a single commercial assay5. In response to these issues, a group of Australian obstetric sonologists formed a collaboration to document the collective national experience of NIPT, across a range of practice types and using a variety of NIPT providers. Australia has approximately 300 000 births per annum and a median maternal age of 30.7 years6. First-trimester combined screening (CFTS), comprising ultrasound measurement of the fetal nuchal translucency and maternal serum pregnancy-associated plasma protein A and beta-human chorionic gonadotropin levels, is currently the dominant form of aneuploidy screening in Australia7. In December 2013, the Australian Association of Obstetrical and Gynaecological Ultrasonologists (AAOGU) invited members to participate in a retrospective audit of NIPT referrals for the period up to the end of 2013, representing the first year of its widespread clinical availability. Participants collected anonymized data on their patients using a standardized data-collection spreadsheet. A combined total of 1839 NIPT referrals, from clinicians practicing in Victoria, New South Wales, Western Australia and Queensland, were made. Most of the 18 practices contributing to the audit contained one or more certified subspecialists in maternal–fetal medicine or obstetric and gynecological ultrasound. The audit included private and public patients accessing NIPT through five different overseas providers based in the USA and China (Harmony Prenatal Test™ (Ariosa Diagnostics, San Jose, CA, USA), Verifi™ (Illumina, San Diego, CA, USA), MaterniT21 Plus™ (Sequenom, San Diego, CA, USA), Panorama™ (Natera, San Carlos, CA, USA) and iGeneScreen™ (BGI, Shenzhen, China)). The Human Research Ethics Committee of Mercy Health gave prospective approval for this study. In the USA, there appeared to be significant uncertainties during the transition of NIPT into clinical care. A 2012 survey of USA maternal–fetal medicine specialists indicated that up to 13% were using NIPT as a diagnostic, rather than a screening, test8. The gestational age at which NIPT was performed in the USA also varied widely, reflecting heterogeneous screening practices9. Australian sonologists who were introduced to NIPT 1 year after their American colleagues had the benefit of several published international consensus statements on NIPT from 2011–201310-12 and a strong paradigm of first-trimester screening. Not surprisingly, in our Australian audit, NIPT was implemented almost exclusively as a first-trimester screen (84.5% at ≤ 14 weeks) in high-risk women (median age, 37 years). Overall, 80% of all tests were performed in the high-risk populations used in the clinical-validity studies, consistent with recommended guidelines at the time11, 13. The two most common indications for NIPT were advanced maternal age or a high-risk CFTS result (Table 1). Within the cohort, two main patterns of implementation could be identified. In the first, NIPT was used as a primary screening test at 10 weeks' gestation, in conjunction with ultrasound examination at 12 weeks, similar to the model reported by Nicolaides et al.4. In the second, NIPT was used as a secondary screen after a high-risk (≥ 1:300) CFTS result, as supported by international consensus statements10, 11 at the time. Importantly, in our experience, NIPT was not used as a replacement for the 11–13-week scan, which was still recommended by the vast majority of sonologists for the purpose of an early fetal structural survey. Less commonly, practitioners justified the continued use of the first-trimester ultrasound examination as a screening test for other adverse obstetric outcomes, such as pre-eclampsia14-16. This retention of the 11–13-week scan in conjunction with first-trimester NIPT has been overlooked in most early cost-effectiveness studies of NIPT17-19 and should be addressed in future policy discussions. All women undergoing NIPT in this audit had evaluation for trisomies 21, 13 and 18; 63% also had full sex-chromosome aneuploidy (SCA) testing. Overall, 45 of the 1839 (2.4%) women received a high-risk NIPT result, of which 27 were for trisomy 21 (Table 2 and Figure 1). Thirty-eight (84.4%) of the women with a high-risk NIPT result intended to have an invasive test, suggesting that clinicians were counseling women appropriately about the need for confirmatory testing. Thirty-five women with a positive NIPT result actually underwent invasive testing (18 amniocentesis and 17 chorionic villus sampling (CVS)), of whom 29 were confirmed to have an abnormal karyotype (positive predictive value (PPV), 82.9%). This PPV reflects the high background risk of the cohort (1.7% overall prevalence of aneuploidy). The total false-positive rate of NIPT was 0.3%, which is consistent with the published literature1. There was one false-positive result for monosomy X that was attributed to low-level maternal mosaicism, illustrating the well-documented capacity for incidental maternal diagnoses with the use of this technology20. If the clinical utility of a prenatal screening test is to be measured by its effect on clinical decision-making, then the Australian NIPT experience has been encouraging. The rate of invasive testing after a low-risk NIPT result was less than 1%, indicating a high confidence in the negative predictive value of NIPT. The downstream impact of NIPT is also reflected in the drastic reduction in prenatal diagnostic procedures over the same period in state-based prenatal diagnostic data21. This reduction is similar to that observed in population-based data on prenatal diagnosis from the USA22. Conversely, there was a high rate of invasive testing in Australian women with high-risk NIPT results, particularly in women at high risk for trisomy 21 (Figure 1). This contrasts with the early USA experience, in which only 33% of women had invasive testing after an abnormal NIPT result9. This suggests that Australian women who chose to undergo NIPT did so with a view to pursuing a definitive prenatal diagnosis if an affected fetus was suspected, against a historical background of high termination rates for fetuses affected by trisomy 21 (> 90%)23. Our rate of invasive testing of 0.9% after a low-risk NIPT result is considerably lower than the published USA9 and UK24 figures of 4.6% and 1.8%, respectively. In the USA study, more women actually had an invasive test after a low-risk NIPT result than did after a high-risk one and almost half of these invasive tests were a result of subsequent screening tests9. The International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) addressed this particular issue in their recent consensus statement on NIPT, recommending that no further aneuploidy risk estimates, including those based on CFTS or the second-trimester 'genetic sonogram', be calculated for a woman who has already received a low-risk NIPT result25. However, the ISUOG statement does emphasize that, in the presence of a fetal structural anomaly (as distinct from a 'soft marker'), the indications for fetal karyotyping should not be modified by a previous normal NIPT result25. In our audit, most of the women who had invasive testing after a low-risk NIPT result did so because of a structural abnormality detected subsequently on ultrasound examination. This figure is very similar to the 0.8% reported by Platt et al. in the USA experience9. One woman in our cohort had a clinically significant atypical chromosomal abnormality (mosaic trisomy 22) ascertained in this manner. This highlights the potential to miss atypical chromosomal abnormalities in high-risk women using the current NIPT panels. The risk of atypical abnormalities in women found to be high risk by CFTS has been estimated to exceed 1.5% in some subgroups26. Our collaboration was able to address several specific considerations related to our geographic location. Over 97% (n = 1787) of women received a result on the first blood draw, with a median turnaround time of 10 calendar days (92% of results received within 14 days). This is similar to the 9-day turnaround time reported in a large industry-sponsored feasibility study from the UK4. The reported 'no result' rate of 2.8% following the first blood draw is also consistent with published figures4, which was reassuring given the relatively long turnaround time, and the low temperatures and pressures associated with international sample shipping. Over 50% of women with no result on the first draw had success on a second draw, but had a median wait of 23 days for a final result. Data are now emerging suggesting that women who experience test failure due to very low fetal fractions may need to be considered as being at increased risk of aneuploidy per se27. In our cohort, four women who had no result by NIPT proceeded to invasive testing and one of these women had a fetus diagnosed with trisomy 21 on CVS. This experience reinforces the need to follow up with some form of additional assessment if NIPT returns no result due to low fetal fraction. This collaboration has provided the first comparison of NIPT uptake between private and public practice in Australia. An AAOGU survey of NIPT users showed that, in December 2013, private practices were significantly more likely to have a specific NIPT policy and to use a single NIPT provider than were public units28. Practitioners identified assay cost as the most important factor in their choice of assay. Reassuringly, there was no significant difference between public and private practices in regards to offering NIPT to high-risk women (95% vs 82%; Fisher's exact test, P = 0.14), suggesting that high-risk women were not disadvantaged by their mode of care in terms of genetic counseling. However, while high-risk women in the public sector may have had equal access to information about NIPT, inequity of access due to cost was identified as a significant ethical issue by more than half of the Australian clinicians offering NIPT. It is revealing that 90% of cases submitted for this audit were from private practices, despite most collaborators working in both the private and public sector. Traditionally, SCA has not been a primary target of prenatal screening, but with its routine availability in NIPT panels, it has rapidly become so. In our audit, 63% of women had full sex-chromosome assessment and 21% had none. The remainder had a limited range of sex-chromosome testing, including diagnosis of fetal sex only (8%). There were 11 cases of suspected SCA, comprising one-quarter of all high-risk NIPT results. All five women with suspected monosomy X underwent invasive testing, but only one of six had testing for other suspected SCAs (Figure 1). This lower rate of invasive testing for SCA compared with the autosomal trisomies has been observed in other clinical studies29. Consequently, the clinical utility of screening for SCA other than monosomy X is not clear, even though women, if asked, do generally elect to receive maximum information30, 31. The newer NIPT panels with selected microdeletions were only introduced into Australia in 2014 and are generally being viewed with caution by clinicians. These tests were not assessed in the audit and it remains to be seen if their yield is perceived to warrant the associated increases in expense, counseling requirements and screen-positive rates32. This study was not designed to assess the sensitivity of NIPT and pregnancy-outcome data were not collected. While one atypical chromosomal abnormality was detected in this dataset following an abnormal ultrasound finding, it is possible that other cases were missed. The fragmentation of NIPT service provision makes any systematic monitoring difficult. However, among our Australian sonologists, there was a high level of support for a future voluntary register of NIPT, incorporating linkage to patient-outcome data and written patient consent for follow-up of pregnancy outcome. Linking NIPT data with prenatal diagnostic procedures, ultrasound screening and pregnancy outcomes is an important priority for future research. In this period of rapid change, government funding to monitor trends in prenatal testing is needed urgently to ensure that the objectives of prenatal screening programs are being met. In December 2013, there was already a high degree of support among Australian sonologists for extending NIPT to low-risk women. Data on the performance of NIPT in unselected or low-risk populations are accumulating4, 27, 33, 34 but it has yet to receive official endorsement from professional societies35. However, before public-health decisions on the applicability of NIPT to the general pregnant population can be made, the current disparity in access must be addressed. There are several mechanisms for doing so. First, commercial pressure to drive down prices has already resulted in substantial cost reductions; the cost of NIPT has more than halved since its arrival in Australia and it is expected to fall further. Second, several countries outside the USA and China are now developing their own local NIPT capabilities, either via technology-transfer arrangements with existing NIPT companies or as completely independent facilities. Third, local cost–benefit analyses using updated prices and implementation models will assist the case for direct government support of this game-changing technology. Our experience of implementing NIPT should resonate with many clinicians, being drawn from a developed country with a universal healthcare system, high uptake of prenatal screening and diagnosis, and patient-funded NIPT. While NIPT has lived up to many of its promises, its appropriate integration with ultrasound examination is still evolving. Regrettably, it was not possible to assess the sensitivity of NIPT in our population due to the lack of monitoring of infrastructure and regulation. In our context, NIPT has changed the way in which antenatal care is delivered in the private sector, but has had only a minimal impact on the public sector. Anticipated reductions in costs and turnaround times with the arrival of onshore providers will be welcomed, as long as excellent test performance can be demonstrated and maintained. With this global shift away from reliance on the USA and Chinese commercial laboratories, trends in prenatal screening and diagnosis and the clinical implementation of NIPT will need to be monitored even more carefully in the coming years. We thank the following individuals who assisted with data collection: Ms Samantha Ayres, Ms Irene Harapa, Ms Joanne Kozary, Ms Linda Lyon, Ms Karen Reidy and Ms Sitarn Ven. In addition to published authors, the following practitioners were members of the NIPT collaboration (listed here in alphabetical order): Dr Michael Bethune, Specialist Women's Ultrasound, Box Hill, VIC; Ms Kellie Chenoweth, Royal Brisbane and Women's Hospital, Herston, QLD; Dr Sophia Chuang, Women's Ultrasound Melbourne, Melbourne, East VIC; Dr Robert Cincotta, Queensland Ultrasound for Women, Spring Hill and Southport, QLD; Dr Stephen Cole, Melbourne Obstetric Group, Epworth Freemasons Hospital, East Melbourne, VIC; Prof Jan Dickinson, Women's Imaging Services, Subiaco, WA; Dr Greg Duncombe, Queensland Ultrasound for Women, Spring Hill and Southport, QLD; Dr Andrew Edwards, City Imaging Ultrasound for Women, East Melbourne, VIC, Camberwell Ultrasound for Women, Hawthorn, East VIC and Central Ultrasound for Women, Fitzroy, VIC; Dr Martha Finn, Monash Ultrasound for Women, Richmond and Clayton, VIC; Dr Beverley Hewitt, Park Ultrasound, West Leederville, WA; Prof Jon Hyett, Royal Prince Alfred Hospital, Camperdown, NSW; Dr Louise Kornman, Women's Ultrasound Melbourne, Melbourne, East VIC; Dr Patricia Lai, Sydney Ultrasound for Women, Newtown, NSW; Dr Valeria Lanzarone, Ultrasound for Women, Penrith, NSW; Dr Anna Lee, Specialist Imaging for Women, Ivanhoe, VIC; Dr Andrew McLennan, Sydney Ultrasound for Women, Newtown, NSW; Dr Simon Meagher, Specialist Imaging for Women, Ivanhoe, VIC; Ms Melody Menezes, Specialist Imaging for Women, Ivanhoe, VIC; Dr Karen Mizia, Ultrasound Care, Newtown, St Leonards, Sydney Adventist Hospital, and Randwick, NSW; Dr Deborah Nesbit, Women's Ultrasound Melbourne, Melbourne, East VIC; Dr Jacqueline Oldham, Women's Ultrasound Melbourne, Melbourne, East VIC; Dr Emily Olive, St Vincent's Private Hospital, Fitzroy, VIC; Dr Alice Robinson, Monash Medical Centre, Clayton, VIC; Dr Steve Raymond, Dr Steve Raymond and Dr Susan Winspear Private Rooms, Lambton, NSW; Dr Shelley Rowlands, Melbourne Obstetric Group, Epworth Freemasons Hospital, East Melbourne, VIC; Dr Amanda Sampson, Women's Ultrasound Melbourne, Melbourne, East VIC; Dr Renuka Sekar, Royal Brisbane and Women's Hospital, Herston, QLD; Dr Kate Stone, Ultrasound for Women, Penrith, NSW; Dr Janet Vaughan, Obstetrics Plus, Willoughby, NSW; Ms Alice Weeks, Specialist Women's Ultrasound, Box Hill, VIC; Dr Susan Winspear, Dr Steve Raymond and Dr Susan Winspear Private Rooms, Fitzroy, VIC; Dr Nicole Woodrow, Women's Ultrasound Melbourne, Melbourne, East VIC.