Prenatal diagnosis routinely is offered to all pregnant women in developed countries who have an increased risk of carrying a child with a chromosomal abnormality. Amniocentesis is the most commonly used invasive prenatal diagnostic procedure worldwide and is performed in 1 out of every 30 pregnancies in developed countries.1,2
Karyotyping detects fetal chromosomal abnormalities in amniotic fluid cells.3,4 It is a robust technique and detects a range of numerical and structural chromosomal abnormalities with high accuracy (99.4–99.9%).3,5,6 However, owing to the required fetal-cell culture, karyotyping is time-consuming and labor-intensive, leading to high costs. The detection capacity of karyotyping may be perceived as a disadvantage because it detects chromosomal abnormalities with unclear or mild clinical relevance. The latter can cause patient anxiety and emotional dilemmas concerning the continuation of pregnancy in situations in which the outcome is uncertain or the phenotype predicted to be relatively mild.7
In the past decade, new molecular techniques have become available for rapid aneuploidy detection of the most common chromosomal abnormalities (aneuploidies of chromosomes X, Y, 13, 18, and 21). Multiplex ligation-dependent probe amplification (MLPA) is a rapid, high-throughput technique shown to be robust in a preclinical setting.8,9 Using MLPA avoids the detection of abnormalities with unclear clinical relevance.
If, under standard clinical conditions, MLPA can detect aneuploidies of chromosomes X, Y, 13, 18, and 21 accurately and rapidly, it would be a suitable test for routine diagnostic application in prenatal diagnosis. Therefore, we conducted a nationwide prospective study in which we compared MLPA with karyotyping in routine clinical practice and evaluated the cost differences of both techniques. We hypothesized that MLPA has equivalent diagnostic accuracy in detecting aneuploidies of chromosomes 21, 13, 18, X, and Y at lower costs.
MATERIALS AND METHODS
The M.A.K.E. (MLPA And Karyotyping, an Evaluation) study was a prospective multicenter diagnostic cohort study comparing MLPA on amniotic fluid in a routine clinical setting with karyotyping.10 All eight Dutch prenatal diagnostic centers and their affiliated hospitals participated. Each institutional review board approved the study, and all participating women gave written informed consent.
We consecutively included pregnant women from March 2007 to October 2008. Pregnant women were eligible for study participation if they had a singleton pregnancy and chose to undergo amniocentesis for advanced maternal age (36 years or older), increased risk of Down syndrome after prenatal screening, or parental anxiety. We excluded women with other indications for amniocentesis (eg, abnormalities on ultrasonography including a nuchal translucency measurement of 3.5 mm, a parental chromosomal abnormality, or a previous child with a chromosomal abnormality) because they have an increased risk of chromosomal abnormalities other than the most common aneuploidies; MLPA cannot detect these, and karyotyping is mandatory.
In all centers, experienced maternal–fetal medicine specialists performed amniocentesis following national guidelines.11 Samples were included if the aspirated volume was at least 14 mL, leaving sufficient amniotic fluid available for MLPA analysis. No extra amniotic fluid was withdrawn for the study.
For the MLPA procedure, DNA was isolated from 1-mL to 8-mL uncultured amniotic fluid samples, depending on the total amount of amniotic fluid received. We used a commercially available kit, the SALSA MLPA P095 (MRC-Holland, Amsterdam, the Netherlands). For each genomic target, a set of two probes is designed to hybridize immediately adjacent to each other on the same target strand. Both probes consist of a short target sequence and a universal polymerase chain reaction (PCR) primer-binding site. One of the probes contains a stuffer sequence with a unique length and sequence. After hybridization, each pair of adjacent probes is joined by a ligation reaction. Next, PCR is performed using a fluorescent-labeled primer pair, which ensures that the relative yield of each of the PCR products is proportional to the amount of each of the target sequences. The different-length products are separated on an automated capillary sequencer. The size and peak areas for each probe are quantified and analyzed by data-analyzing software (GeneMarker, SoftGenetics, LLC, State College, PA, or Genescan and Genemapper 3.7/4.0, Applied Biosystems, Foster City, CA).8 Relative probe signals are calculated and compared with samples from chromosomally normal male and female fetuses. In chromosomally normal samples, the relative probe signal is expected to be 1 for all probes. A normal value is defined as a relative probe signal between 0.7 and 1.3. A relative probe value of less than 0.7 indicates a monosomy, and a relative probe value of more than 1.3 indicates a trisomy. It is not expected for MLPA to detect low-grade chromosomal mosaicism.9,12
Technicians had a molecular genetics or a cytogenetics background; all were trained in the execution of MLPA before the start of the study. Provided that at least 2 mL of amniotic fluid was available, MLPA was performed in duplicate. Results of MLPA were conclusive if both tests matched. If the results one or either test were inconclusive and sufficient DNA was available, the MLPA reaction was repeated. If the results still disagreed after the repetition, MLPA failed. Technicians carrying out MLPA were blinded to karyotyping results and vice versa. However, if MLPA detected an aneuploidy, the head of the laboratory could initiate the earliest possible harvesting of cell culture.
We allowed a phase 1 (median time 6 months) in which test results were not reported to patients and centers could train extra personnel for sample identification, tracking, and accurate reporting of test results. In phase 2, conclusive MLPA results were reported to pregnant women as a provisional result while awaiting the definite karyotype result. Patients also were informed if MLPA failed. For karyotyping, fetal cells were cultured and spread on slides, which were stained for chromosomal banding. Routinely, metaphases for 10 colonies were investigated. All centers followed national quality guidelines, but minor differences in the amount of cell colonies cultured, staining, and reporting of the results were allowed.13
The primary outcome variable was diagnostic accuracy for detecting aneuploidies of chromosomes 21, 13, 18, X, and Y. We quantified the other chromosomal abnormalities that were not detected by MLPA and recorded reasons for failed test results. Turnaround time for test results was measured at the laboratory level (time span between carrying out the amniocentesis and authorization of test result) and, in phase 2, at the patient level (time span between amniocentesis and the result given to the patient).
Mean cost differences between MLPA and karyotyping as standalone strategies were evaluated according to international guidelines.14,15 Costs per strategy were calculated as the sum of resource use between amniocentesis and the decision to continue or terminate pregnancy; individual data from the case-record forms and direct observations in three centers were multiplied by resource unit prices covering personnel costs, equipment, consumables, additional costs in case of chromosomal abnormality, and overhead costs. Costs were calculated in Euros and then converted into U.S. dollars (€1.00=U.S. $1.37).
Sample size was estimated to demonstrate noninferiority of the index test (MLPA) to karyotyping. During a pretrial meeting, experts in prenatal diagnosis, clinical epidemiology, and statistics agreed on a critical noninferiority margin of 0.002. At least 4,497 paired test results were needed (one-sided alpha 0.05, power 0.90) to reject the null hypothesis that MLPA is inferior to karyotyping. We calculated diagnostic accuracy by dividing the sum of the true-positive and true-negative results by the total number of participants. Sensitivity and specificity were calculated by standard formulas for binomial proportions; 95% confidence intervals (CIs) were calculated by the Wilson interval method.16,17 Failed results were expressed in absolute numbers and percentages. To identify patient, procedural, and center-specific characteristics associated with failure rate, we performed backward-selection logistic regression analysis. Differences in costs were tested with the Student t tests (SPSS 16.0, SPSS, Inc., Chicago, IL). Differences in turnaround time for test results were compared with a Kruskal-Wallis test followed by the post hoc Dunn's test.
In total 4,649 women were eligible and 64 (1.4%) were excluded; 4,585 amniotic fluid samples were tested with both MLPA and karyotyping (Fig. 1). The laboratory results for 280 of the women were published previously.18 Patient and procedural characteristics are listed in Tables 1 and 2. In 4,484 of the 4,585 samples (97.8%), MLPA and karyotyping were concordant, showing normal results in 4,386 of 4,585 (95.7%) and aneuploidy in 98 of 4,585 (2.1%) (Table 3). Discordant results were found in 26 of 4,585 (0.6%) samples, representing an abnormal karyotype not detected by MLPA (Table 4 and Box 1 [available online at http://links.lww.com/AOG/A152]). Diagnostic accuracy of MLPA was 1.0 (95% CI 0.99–1.0) with a sensitivity of 100% (95% CI 0.96–1.0) and a specificity of 100% (95% CI 0.999–1.0). Therefore, we rejected the null hypothesis that MLPA is inferior to karyotyping (P<.001).
In 75 cases (1.6%), the MLPA test failed. Karyotyping failed in 1 of these 75 cases (0.02%). The failure rate of MLPA was 2.4% in the first 4 months of the study, decreasing to 1.5% in the last 11 months. Variables significantly associated with increasing failure rate were contaminated amniotic fluid (odds ratio [OR] 5.29, 95% CI 2.4–11.6) and contaminated cell pellet (OR 3.39, 95% CI 1.98–5.81). Variables significantly associated with a lower risk of failure were time from start of study participation (per month OR 0.95, 95% CI 0.90–0.99) and milliliters of amniotic fluid available for MLPA (per mL OR 0.78, 95% CI 0.69–0.88). Compared with a dual-operator technique, the single-operator technique with (OR 0.22, 95% CI 0.1–0.48) and without (OR 0.28, 95% CI 0.14–0.55) continuous ultrasound control was significantly associated with a lower risk of failure.
We performed 1,223 MLPA reactions in phase 1 (median time for phase 1 was 6 months) and 3,362 in phase 2. Median laboratory turnaround time for MLPA was 6 days (interquartile range 4–8) in phase 1 and 3 days (interquartile range 2–7) in phase 2; median laboratory turnaround time was 17 days (interquartile range 15–20) for karyotyping (Fig. 2) (medians for phase 1 compared with phase 2 compared with karyotyping: P<.001; medians for phase 1 compared with phase 2: P<.001; medians for phase 1 compared with karyotyping: P<.001; medians for phase 2 compared with karyotyping: P<.001; all Kruskal-Wallis test followed by Dunn's test).
Median time between amniocentesis and informing pregnant women was 3 days (interquartile range 3–7) for MLPA and 18 days (interquartile range 16–21) for karyotyping. Mean time reduction of MLPA compared with karyotyping was 13.8 days (P<.001, 95% CI 13.7–14.0) and 14.5 days (P<.001, 95% CI 14.3–14.6) at the laboratory and patient level, respectively.
Costs for MLPA were $472.00. Costs for karyotyping were $915.00. Mean cost reduction per sample was $433 (95% CI $416–$449; −47%) in favor of MLPA (P<.001).
In this nationwide prospective cohort study including more than 4,500 women, we demonstrated that the diagnostic accuracy of MLPA to detect aneuploidies of chromosomes 21, 13, 18, X, and Y is comparable with that of karyotyping and that MLPA is less costly than karyotyping. Our large study under standard practice conditions confirms and extends the findings of recent preclinical studies on MLPA.19,20 Compared with other techniques for rapid aneuploidy detection, the diagnostic accuracy of MLPA is similar to that of quantitative fluorescent PCR (0.99–1.0) and fluorescence in situ hybridization (FISH) (0.99–1.0), with comparable failure rates of 0.1–3.7% for quantitative fluorescent PCR and 0.0–4.9% for FISH.21–26 However, few of these results were obtained under practice conditions. Compared with quantitative fluorescent PCR, MLPA is relatively sensitive to DNA quality and does not detect maternal cell contamination in female samples or female triploidies. Forty genomic targets can be detected by MLPA in one reaction, and MLPA avoids the problem of noninformativeness of the polymorphic markers that may occur with quantitative fluorescent PCR. Compared with FISH, MLPA and quantitative fluorescent PCR are both more suitable for high-throughput testing at lower costs.22 Therefore, quantitative fluorescent PCR and MLPA represent the preferred techniques for routine prenatal diagnosis. If chromosomal mosaicism is suspected, however, FISH is preferred because detection levels of 5% can be achieved.23
Our study showed that MLPA has lower costs compared with karyotyping; however, similar to studies on quantitative fluorescent PCR and FISH, considerable variation among laboratories exists, mainly caused by differences in sample throughput and logistics.22 Further research is warranted to determine the additional costs accrued by lifetime costs of chromosomal abnormalities. The failure rate of 1.6%, similar to previous studies,12,19,20 is a concern. In a standalone policy, failure implies repeating the amniocentesis with its inherent risks. It is likely that the true failure rate in a standalone policy is lower. Firstly, there was a 38% reduction of the failure rate (from 2.4% to 1.5%) between early and later experience with the test. Secondly, the study protocol prioritized karyotyping, which requires 12 mL of amniotic fluid. In a standalone policy, more amniotic fluid is available for MLPA and the failure rate will fall. Thirdly, a further decrease in the failure rate may occur when a lower number of bloody samples can be achieved. The American College of Obstetricians and Gynecologists recommends continuously visualizing the needle for this purpose.27 From our study results and the available evidence, we recommend using the single-operator technique with continuous ultrasound control. Furthermore, there are two options to manage macroscopically blood-stained samples. One option is to detect the proportion of fetal hemoglobin compared with adult hemoglobin and to perform MLPA if the fetal hemoglobin level is 85% or more of the total hemoglobin level.20 The other option is to omit MLPA and perform karyotyping on these samples. Finally, in a standalone policy, we recommend short-term storage of amniotic fluid cells to allow karyotyping should MLPA fail and subsequent storage of DNA to allow follow-up molecular diagnostics without repeated amniocentesis should ultrasound examination show an abnormality.
The main argument against replacing karyotyping with rapid aneuploidy detection is that some clinically severe chromosomal abnormalities will remain undetected. Of the 26 chromosomal abnormalities (out of 4,585; 0.6%) that MLPA could not detect, 17 were without clinical consequences for the current pregnancy (Box 1). Of these, 14 were inherited balanced rearrangements, which may lead to future unbalanced rearrangements. Six of the remaining nine abnormalities were chromosomal abnormalities with uncertain clinical consequences. If detected, this type of abnormality leads to difficult counseling issues and emotional dilemmas.7 It is questionable whether their detection is in the best interest of the parents because it may lead to an unwarranted termination of pregnancy.1,28 The last three chromosomal abnormalities were of serious clinical significance (Box 1); this overall residual risk of 0.07% confirms findings by others.1,21 In our study, with knowledge of the karyotype, standard follow-up ultrasound examination showed abnormalities in one out of the three. Hence, when using standalone MLPA combined with ultrasound examination, two chromosomal abnormalities of serious clinical significance remain undetected. In total, 3 of the 26 pregnancies were terminated (one of uncertain clinical consequence and two of serious clinical significance) and 23 were continued. Therefore, in our sample of 4,585 pregnancies, the added knowledge from karyotyping lead to three extra terminations of pregnancy.
The provision of rapid, unambiguous, and low-cost results is an incentive to implement MLPA. Successful implementation also requires the support of pregnant women.29 So far, two studies show that pregnant women prefer rapid aneuploidy detection over karyotyping.30,22 A Swedish study showed that 70% of women who were offered an actual choice preferred rapid testing over karyotyping.31 At the public-health level, these studies suggest that rapid testing is the preferred strategy. If one adheres to individual choice, it could be argued that the decision to either obtain as much cytogenetic information as possible or to receive a rapid, specific result is most appropriately made by the individuals who will bear the responsibility of raising the child.
In this era of rapid developments in prenatal diagnosis, the debate about what to test for remains essential. At present, the use of microarrays, which can detect even more chromosomal abnormalities than karyotyping can, is being studied.32 Within a few years, noninvasive diagnosis of fetal chromosomal abnormalities in maternal blood may be available,33 eliminating the procedure-related miscarriage risk. Even with these new developments, the debate about targeted- or whole-genome testing remains in force. The widespread introduction of molecular tests changes the scope of prenatal diagnosis and should encourage the development of strategies that tailor the type of diagnostic test offered to the risk identified. Future studies should focus on the application of tailor-made strategies, including the views of pregnant women and possible barriers that hamper successful implementation of new prenatal test strategies. For now, the use of MLPA in prenatal diagnosis appears to be a prudent strategy.
1.Ogilvie CM, Lashwood A, Chitty L, Waters JJ, Scriven PN, Flinter F. The future of prenatal diagnosis: rapid testing or full karyotype? An audit of chromosome abnormalities and pregnancy outcomes for women referred for Down's Syndrome testing. BJOG 2005;112:1369–75.
3.Midtrimester amniocentesis for prenatal diagnosis. Safety and accuracy. JAMA 1976;236:1471–6.
4.Tabor A, Madsen M, Obel EB, Philip J, Bang J, Norgaard-Pedersen B. Randomised controlled trial of genetic amniocentesis in 4606 low-risk women. Lancet 1986;1:1287–93.
5.Los FJ, van Den Berg C, Wildschut HI, Brandenburg H, den Hollander NS, Schoonderwaldt EM, et al. The diagnostic performance of cytogenetic investigation in amniotic fluid cells and chorionic villi. Pren Diagn 2001;21:150–8.
6.Golbus MS, Loughman WD, Epstein CJ, Halbasch G, Stephens JD, Hall BD. Prenatal genetic diagnosis in 3000 amniocentesis. N Engl J Med 1979;300:157–63.
7.van Zwieten MCB, Willems DL, Litjens LL, Schuring-Blom HG, Leschot N. How unexpected are unexpected findings in prenatal cytogenetic diagnosis? A literature review. Eur J Obstet Gynecol Reprod Biol 2005;120:15–20.
8.Slater HR, Bruno DL, Ren H, Pertile M, Schouten JP, Choo KH. Rapid, high throughput prenatal detection of aneuploidy using a novel quantitative method (MLPA). J Med Genet 2003;40:907–12.
9.Gerdes T, Kirchhoff M, Lind AM, Vestergaard Larsen G, Schwartz M, Lundsteen C. Computer-assisted prenatal aneuploidy screening for chromosome 13, 18, 21, X and Y based on multiplex ligation-dependent probe amplification (MLPA). Eur J Hum Gen 2005;13:171–5.
10.Boormans EM, Birnie E, Wildschut HI, Schuring-Blom HG, Oepkes D, van Oppen CA, et al. Multiplex ligation-dependent probe amplification versus karyotyping in prenatal diagnosis: the MA.K.E. study. BMC Pregnancy Childbirth 2008;8:18.
11.Dutch Society of Obstetrics and Gynecology. Quality guideline invasive prenatal diagnosis. Utrecht (Netherlands): NVOG; 1997.
12.Hochstenbach R, Meijer J, van de Brug J, Vossebeld-Hoff I, Jansen R, van der Luijt RB, et al. Rapid detection of chromosomal aneuploidies in uncultured amniocytes by multiplex ligation-dependent probe amplification (MLPA). Prenat Diagn 2005;25:1032–9.
13.Dutch Society of Human Genetics. Quality in clinical cytogenetics: conditions, standards and tests. Amsterdam (Netherlands): NVHG; 2003.
14.Oostenbrink JB, Boumans CAM, Koopmanschap MA, Rutten FFH. Handleiding voor kostenonderzoek, methoden en standaard kostprijzen voor economische evaluaties in de gezondheidszorg. Diemen (Netherlands): College voor zorgverzekeringen; 2004.
15.Gold MR, Siegel JE, Russell LB, Weinstein MC. Cost-effectiveness in health and medicine. New York (NY): Oxford University Press; 1996.
16.Agresti A, Coull BA. Approximate is better than “exact” for interval estimation of binomial proportions. Am Stat 1998;52:119–26.
17.Newcombe RG. Two-sided confidence intervals for the single proportion: comparison of seven methods. Stat Med 1998;17:857–72.
18.Kooper AJA, Faas BHW, Kater-Baats E, Feuth T, Janssen JC, van der Burgt I, et al. Multiplex ligation-dependent probe amplification (MLPA) as a stand-alone test for rapid aneuploidy detection in amniotic fluid cells. Pren Diagn 2008;28:1004–10.
19.Gerdes T, Kirchhoff M, Lind AM, Vestergaard Larsen G, Kjaergaard S. Multiplex ligation-dependent probe amplification (MLPA) in prenatal diagnosis-experience of a large series of rapid testing for aneuploidy of chromosomes 13, 18, 21, X and Y. Pren Diagn 2008;28:1119–25.
20.van Opstal D, Boter M, de Jong D, van den Berg C, Brüggenwirth HT, Wildschut HIJ, et al. Rapid aneuploidy detection with multiplex ligation-dependent probe amplification: a prospective study of 4000 amniotic fluid samples. Eur J Hum Gen 2009;17:112–21.
21.Shaffer LG, Bui TH. Molecular cytogenetic and rapid aneuploidy detection methods in prenatal diagnosis. Am J Med Genet C Semin Med Genet 2007;145C:87–98.
22.Grimshaw GM, Szczepura A, Hultén M, MacDonald F, Nevin NC, Sutton F, et al. Evaluation of molecular tests for prenatal diagnosis of chromosome abnormalities. Health Technol Assess 2003;7:1–146.
23.Mann K, Fox SP, Abbs SJ, Ching Yau S, Scriven PN, Docherty Z, et al. Development and implementation of a new rapid aneuploidy diagnostic service within the UK National Health service and implications for the future of prenatal diagnosis. Lancet 2001;358:1057–61.
24.Tepperberg J, Pettenati MJ, Rao PN, Lese CM, Rita D, Wyandt H, et al. Prenatal diagnosis using interphase fluorescence in situ hybridization (FISH): 2-year multi-center retrospective study and review of the literature. Pren Diagn 2001;21:293–301.
25.Ogilvie CM, Donaghue C, Fox SP, Docherty Z, Mann K. Rapid prenatal diagnosis of aneuploidy using quantitative fluorescence-PCR (QF-PCR). J of Histochem Cytochem 2005;53:285–8.
26.Cirigliano V, Voglino G, Ordoñez E, Marongiu A, Paz Cañadas M, Ejarque M, et al. Rapid prenatal diagnosis of common chromosome aneuploidies by QF-PCR, results of 9 years of clinical experience. Pren Diagn 2009;29:40–9.
27.Invasive prenatal testing for aneuploidy. ACOG Practice Bulletin No. 88. American College of Obstetricians and Gynecologists. Obstet Gynecol 2007;110:1459–67.
28.Leung WC, Lau ET, Lao TT, Tang MH. Rapid aneuploidy screening (FISH or QF-PCR): the changing scene in prenatal diagnosis? Expert Rev Mol Diagn 2004;4:333–7.
29.Kassirer JP. Incorporating patients' preferences into medical decisions. N Engl J Med 1994;330:1895–6.
30.Ryan M, Diack J, Watson V, Smith N. Rapid prenatal diagnostic testing for Down syndrome only or longer wait for full karyotype: the views of pregnant women. Pren Diagn 2005;25:1206–11.
31.Bui TH. Prenatal cytogenetic diagnosis: gone FISHing, BAC soon! Ultrasound Obstet Gynecol 2007;30:247–51.
32.Van den Veyver IB, Beaudet AL. Comparative genomic hybridization and prenatal diagnosis. Curr Opin Obstet Gynecol 2006;18:185–91.
33.Lo YM. Fetal nucleic acids in maternal plasma. Ann N Y Acad Sci 2008;1137:140–3.
Supplemental Digital Content
© 2010 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.