Analytical Validation of a Deep Neural Network Algorithm for the Detection of Ovarian Cancer.

PURPOSE: Early detection of ovarian cancer, the deadliest gynecologic cancer, is crucial for reducing mortality. Current noninvasive risk assessment measures include protein biomarkers in combination with other clinical factors, which vary in their accuracy. Machine learning can be applied to optimizing the combination of these features, leading to more accurate assessment of malignancy. However, the low prevalence of the disease can make rigorous validation of these tests challenging and can result in unbalanced performance. METHODS: MIA3G is a deep feedforward neural network for ovarian cancer risk assessment, using seven protein biomarkers along with age and menopausal status as input features. The algorithm was developed on a heterogenous data set of 1,067 serum specimens from women with adnexal masses (prevalence = 31.8%). It was subsequently validated on a cohort almost twice that size (N = 2,000). RESULTS: In the analytical validation data set (prevalence = 4.9%), MIA3G demonstrated a sensitivity of 89.8% and a specificity of 84.02%. The positive predictive value was 22.45%, and the negative predictive value was 99.38%. When stratified by cancer type and stage, MIA3G achieved sensitivities of 94.94% for epithelial ovarian cancer, 76.92% for early-stage cancer, and 98.04% for late-stage cancer. CONCLUSION: The balanced performance of MIA3G leads to a high sensitivity and high specificity, a combination that may be clinically useful for providers in evaluating the appropriate management strategy for their patients. Limitations of this work include the largely retrospective nature of the data set and the unequal, albeit random, assignment of histologic subtypes between the training and validation data sets. Future directions may include the addition of new biomarkers or other modalities to strengthen the performance of the algorithm.

Salvaging Detection of Early-Stage Ovarian Malignancies When CA125 Is Not Informative.

BACKGROUND: Ovarian cancer is the deadliest gynecologic cancer, with no recommended screening test to assist with early detection. Cancer antigen 125 (CA125) is a serum biomarker commonly used by clinicians to assess preoperative cancer risk, but it underperforms in premenopausal women, early-stage malignancies, and several histologic subtypes. OVA1 is a multivariate index assay that combines CA125 and four other serum proteins to assess the malignant risk of an adnexal mass. OBJECTIVE: To evaluate the performance of OVA1 in a cohort of patients with low-risk serum CA125 values. STUDY DESIGN: We analyzed patient data from previous collections (N = 2305, prevalence = 4.5%) where CA125 levels were at or below 67 units/milliliter (U/mL) for pre-menopausal women and 35 U/mL for post-menopausal women. We compare the performance of OVA1 to CA125 in classifying the risk of malignancy in this cohort, including sensitivity, specificity, positive and negative predictive values. RESULTS: The overall sensitivity of OVA1 in patients with a low-risk serum CA125 was 59% with a false-positive rate of 30%. OVA1 detected over 50% of ovarian malignancies in premenopausal women despite a low-risk serum CA125. OVA1 also correctly identified 63% of early-stage cancers missed by CA125. The most common epithelial ovarian cancer subtypes in the study population were mucinous (25%) and serous (23%) carcinomas. Despite a low-risk CA125, OVA1 successfully detected 83% of serous, 58% of mucinous, and 50% of clear cell ovarian cancers. CONCLUSIONS: As a standalone test, CA125 misses a significant number of ovarian malignancies that can be detected by OVA1. This is particularly important for premenopausal women and early-stage cancers, which have a much better long-term survival than late-stage malignancies. Using OVA1 in the setting of a normal serum CA125 can help identify at-risk ovarian tumors for referral to a gynecologic oncologist, potentially improving overall survival.

Development and validation of concurrent preimplantation genetic diagnosis for single gene disorders and comprehensive chromosomal aneuploidy screening without whole genome amplification.

OBJECTIVE: To develop a novel and robust protocol for multifactorial preimplantation genetic testing of trophectoderm biopsies using quantitative polymerase chain reaction (qPCR). DESIGN: Prospective and blinded. SETTING: Not applicable. PATIENT(S): Couples indicated for preimplantation genetic diagnosis (PGD). INTERVENTION(S): None. MAIN OUTCOME MEASURE(S): Allele dropout (ADO) and failed amplification rate, genotyping consistency, chromosome screening success rate, and clinical outcomes of qPCR-based screening. RESULT(S): The ADO frequency on a single cell from a fibroblast cell line was 1.64% (18/1,096). When two or more cells were tested, the ADO frequency dropped to 0.02% (1/4,426). The rate of amplification failure was 1.38% (55/4,000) overall, with 2.5% (20/800) for single cells and 1.09% (35/3,200) for samples that had two or more cells. Among 152 embryos tested in 17 cases by qPCR-based PGD and CCS, 100% were successfully given a diagnosis, with 0% ADO or amplification failure. Genotyping consistency with reference laboratory results was >99%. Another 304 embryos from 43 cases were included in the clinical application of qPCR-based PGD and CCS, for which 99.7% (303/304) of the embryos were given a definitive diagnosis, with only 0.3% (1/304) having an inconclusive result owing to recombination. In patients receiving a transfer with follow-up, the pregnancy rate was 82% (27/33). CONCLUSION(S): This study demonstrates that the use of qPCR for PGD testing delivers consistent and more reliable results than existing methods and that single gene disorder PGD can be run concurrently with CCS without the need for additional embryo biopsy or whole genome amplification.

Nuclear localization of SMN and FUS is not altered in fibroblasts from patients with sporadic ALS.

Abstract Sporadic amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease with no established biological marker. Recent observation of a reduced number of gems (survival motor neuron protein (SMN)-positive nuclear bodies) in cells from patients with familial ALS and the mouse models suggests an involvement of SMN in ALS pathology. At a molecular level, fused in sarcoma (FUS), one of the familial ALS-linked proteins, has been demonstrated to directly interact with SMN, while impaired nuclear localization of mutated FUS causes defective gem formation. Our objective was to determine whether gems and/or nuclear FUS levels in skin derived fibroblasts from sporadic ALS patients are consistently reduced and thus could constitute a novel and readily available biomarker of the disease. Fibroblasts from 20 patients and 17 age-matched healthy controls were cultured and co-immunostained for SMN and FUS. Results showed that no difference was detected between the two groups in the number of gems and in expression pattern of FUS. The number of gems negatively correlated with the age at biopsy in both ALS and control subjects. In conclusion, the expression pattern of SMN and FUS in fibroblasts cannot serve as a biomarker for sporadic ALS. Donor age-dependent gem reduction is a novel observation that links SMN with cellular senescence.

Tracking embryo implantation using cell-free fetal DNA enriched from maternal circulation at 9 weeks gestation.

Embryo DNA fingerprinting represents an important tool for tracking embryo-specific outcomes after multiple embryo transfer during IVF. The situation in which two embryos are transferred and only one implants represents a unique opportunity for the most well-controlled validation of markers capable of identifying competent and incompetent embryos. Specifically, this design eliminates all patient-related variables from the comparison of embryos with or without reproductive potential. However, in order to determine which embryo implanted, the investigator must wait until newborn DNA is available upon delivery. This study validates a non-invasive cell-free fetal DNA fingerprinting method that reduces the time to identify which embryo implanted by ∼31 weeks. Thirty-four patients were studied to determine if fingerprinting of fetal DNA extracted from maternal plasma at nine gestational weeks concurred with the buccal DNA results obtained from the newborn after delivery. This validation required single nucleotide polymorphism profiles on each couples' preimplantation embryos, cell-free fetal DNA from maternal plasma at 9 weeks gestation, and newborn DNA obtained from buccal swabs after delivery. The predictions from fetal DNA-based embryo tracking and gender assignments made at 9 weeks gestation were 100% consistent with standardized methods of assessment performed after term delivery. This study demonstrates the first validated fetal DNA fingerprinting method which predicts both gender and which embryo implanted at 9 weeks gestation following multiple embryo transfer.

Single nucleotide polymorphism microarray-based concurrent screening of 24-chromosome aneuploidy and unbalanced translocations in preimplantation human embryos.

OBJECTIVE: To develop, validate, and apply a single nucleotide polymorphism (SNP) microarray-based method for simultaneous preimplantation genetic diagnosis (PGD) of unbalanced inheritance of rearranged chromosomes and 24-chromosome aneuploidy screening. DESIGN: Prospective clinical research study. SETTING: Academic reproductive medicine center. PATIENT(S): Eighteen couples carrying a balanced reciprocal or Robertsonian chromosomal rearrangement. INTERVENTION(S): PGD on blastocyst trophectoderm biopsy specimens. MAIN OUTCOME MEASURE(S): Aneuploidy, implantation, pregnancy, and delivery rates after SNP microarray-based aneuploidy and translocation screening. RESULT(S): Single nucleotide polymorphism microarray was capable of detecting translocation-associated imbalances as small as 9.0 megabases. In the 12 transfers performed, sustained implantation occurred for 9 (45%) of 20 balanced-normal and euploid embryos replaced. The clinical pregnancy rate in patients receiving a transfer was 75% with six singleton deliveries and three ongoing singleton pregnancies thus far. Significantly fewer embryos were eligible for transfer with the incorporation of simultaneous 24-chromosome aneuploidy screening. Arrested embryos were also significantly more likely to possess unbalanced chromosomes when compared with developmentally competent blastocysts. CONCLUSION(S): This SNP microarray-based method provides the first opportunity to improve outcomes through comprehensive identification of euploid embryos from translocation carrier couples.

Single-cell whole-genome amplification technique impacts the accuracy of SNP microarray-based genotyping and copy number analyses.

Methods of comprehensive microarray-based aneuploidy screening in single cells are rapidly emerging. Whole-genome amplification (WGA) remains a critical component for these methods to be successful. A number of commercially available WGA kits have been independently utilized in previous single-cell microarray studies. However, direct comparison of their performance on single cells has not been conducted. The present study demonstrates that among previously published methods, a single-cell GenomePlex WGA protocol provides the best combination of speed and accuracy for single nucleotide polymorphism microarray-based copy number (CN) analysis when compared with a REPLI-g- or GenomiPhi-based protocol. Alternatively, for applications that do not have constraints on turnaround time and that are directed at accurate genotyping rather than CN assignments, a REPLI-g-based protocol may provide the best solution.

SNP microarray-based 24 chromosome aneuploidy screening is significantly more consistent than FISH.

Many studies estimate that chromosomal mosaicism within the cleavage-stage human embryo is high. However, comparison of two unique methods of aneuploidy screening of blastomeres within the same embryo has not been conducted and may indicate whether mosaicism has been overestimated due to technical inconsistency rather than the biological phenomena. The present study investigates the prevalence of chromosomal abnormality and mosaicism found with two different single cell aneuploidy screening techniques. Thirteen arrested cleavage-stage embryos were studied. Each was biopsied into individual cells (n = 160). The cells from each embryo were randomized into two groups. Those destined for FISH-based aneuploidy screening (n = 75) were fixed, one cell per slide. Cells for SNP microarray-based aneuploidy screening (n = 85) were put into individual tubes. Microarray was significantly more reliable (96%) than FISH (83%) for providing an interpretable result (P = 0.004). Markedly different results were obtained when comparing microarray and FISH results from individual embryos. Mosaicism was significantly less commonly observed by microarray (31%) than by FISH (100%) (P = 0.0005). Although FISH evaluated fewer chromosomes per cell and fewer cells per embryo, FISH still displayed significantly more unique genetic diagnoses per embryo (3.2 +/- 0.2) than microarray (1.3 +/- 0.2) (P < 0.0001). This is the first prospective, randomized, blinded and paired comparison between microarray and FISH-based aneuploidy screening. SNP microarray-based 24 chromosome aneuploidy screening provides more complete and consistent results than FISH. These results also suggest that FISH technology may overestimate the contribution of mitotic error to the origin of aneuploidy at the cleavage stage of human embryogenesis.

Selective oxytocin receptor activation in the ventrolateral portion of the ventromedial hypothalamus is required for mating-induced pseudopregnancy in the female rat.

The ventrolateral region of the ventromedial hypothalamus (VMHvl) plays an essential role in female sexual behavior. Oxytocin (OT) is released from the paraventricular nucleus to downstream sites such as the VMHvl to facilitate female sexual behavior and shows characteristics of a prolactin (PRL)-releasing factor. During mating, vaginal cervical stimulation (VCS) received from a vasectomized male triggers twice-daily PRL surges that persist up to 12+ d, a period known as pseudopregnancy (PSP). To determine whether OT is involved in PSP by acting within the VMHvl, female rats were infused bilaterally with an oxytocin receptor antagonist (OTR-A), a vasopressin receptor-1a antagonist (V(1a)-A), or artificial cerebral spinal fluid 30 min before mating. All females received a sufficient amount of VCS, 15 intromissions, to induce PSP. Females infused with OTR-A (20 ng/0.4 microl) with implants targeting the VMHvl showed only a 22% induction of PSP, as measured using vaginal diestrus and serum PRL concentrations. In contrast, controls and V(1a)-A (80 ng/0.4 microl) infused females exhibited 100% induction of PSP. Females infused with OTR-A returned to estrus after 5 d, whereas females infused with either artificial cerebral spinal fluid or V(1a)-A remained in diestrus for 12-13 d in both the correct and missed placement groups. Although OT can act as a PRL releasing factor, the PRL surge does not begin until 18-24 h after mating. Together, our results suggest that OT release in the VMHvl mediates the effects of VCS on the induction of the PRL secretion needed to establish PSP.