Multilevel regression modeling for aneuploidy classification and physical separation of maternal cell contamination facilitates the QF-PCR based analysis of common fetal aneuploidies

Background The quantitative fluorescent polymerase chain reaction (QF-PCR) has proven to be a reliable method for detection of common fetal chromosomal aneuploidies. However, there are some technical shortcomings, such as uncertainty of aneuploidy determination when the short tandem repeats (STR) height ratio is unusual due to a large size difference between alleles or failure due to the presence of maternal cell contamination (MCC). The aim of our study is to facilitate the implementation of the QF-PCR as a rapid diagnostic test for common fetal aneuploidies. Methods Here, we describe an in-house one-tube multiplex QF-PCR method including 20 PCR markers (15 STR markers and 5 fixed size) for rapid prenatal diagnosis of chromosome 13, 18, 21, X and Y aneuploidies. In order to improve the aneuploidy classification of a given diallelic STR marker, we have employed a multilevel logistic regression analysis using height-ratio and allele-size-difference as fixed effects and marker as a random effect. We employed two regression models, one for the 2:1 height ratio (n = 48 genotypes) and another for the 1:2 height ratio (n = 41 genotypes) of the trisomic diallelic markers while using the same 9015 genotypes with normal 1:1 height ratio in both models. Furthermore, we have described a simple procedure for the treatment of the MCC, prior DNA isolation and QF-PCR analysis. Results For both models, we have achieved 100% specificity for the marker aneuploidy classification as compared to 98.60% (2:1 ratio) and 98.04% (1:2 ratio) specificity when using only the height ratio for classification. Treatment of the MCC enables a successful diagnosis rate of 76% among truly contaminated amniotic fluids. Conclusions Adjustment for the allele size difference and marker type improves the STR aneuploidy classification, which, complemented with appropriate treatment of contaminated amniotic fluids, eliminates sample re-testing and reinforces the robustness of the QF-PCR method for prenatal testing.