Evaluation of machine learning-driven automated Kleihauer-Betke counting: A method comparison study.

INTRODUCTION: The Kleihauer-Betke (KB) test is the diagnostic standard for the quantification of fetomaternal hemorrhage (FMH). Manual analysis of KB slides suffers from inter-observer and inter-laboratory variability and low efficiency. Flow cytometry provides accurate quantification of FMH with high efficiency but is not available in all hospitals or at all times. We have developed an automated KB counting system that uses machine learning to identify and distinguish fetal and maternal red blood cells (RBCs). In this study, we aimed to evaluate and compare the accuracy, precision, and efficiency of the automated KB counting system with manual KB counting and flow cytometry. METHODS: The ratio of fetal RBCs of the same blood sample was quantified by manual KB counting, automated KB counting, and flow cytometry, respectively. Forty patients were enrolled in this comparison study. RESULTS: Comparing the automated KB counting system with flow cytometry, the mean bias in measuring the ratio of fetal RBCs was 0.0048%, with limits of agreement ranging from -0.22% to 0.23%. Using flow cytometry results as a benchmark, results of automated KB counting were more accurate than those from manual counting, with a lower mean bias and narrower limits of agreement. The precision of automated KB counting was higher than that of manual KB counting (intraclass correlation coefficient 0.996 vs 0.79). The efficiency of automated KB counting was 200 times that of manual counting by the certified technologists. CONCLUSION: Automated KB counting provides accurate and precise FMH quantification results with high efficiency.

Evaluating emergency-release blood transfusion of newborn infants at the Intermountain Healthcare hospitals.

BACKGROUND: An emergency-release blood transfusion (ERBT) protocol (uncrossmatched type O-negative red blood cells, AB plasma, AB platelets) is critical for neonatology practice. However, few reports of emergency transfusions are available. We conducted an ERBT quality improvement project as a basis for progress. STUDY DESIGN AND METHODS: For each ERBT in the past 8 years, we logged indications, products, locations and timing of the transfusions, and outcomes. RESULTS: One hundred forty-nine ERBTs were administered; 42% involved a single blood product, and 58% involved two or more. The incidence was 6.25 ERBT per 10,000 live births, with a higher rate (9.52 ERBT/10,000) in hospitals with a Level 3 neonatal intensive care unit (NICU) (p < 0.001). Seventy percent of ERBTs were administered in a NICU and 30% in a delivery room, operating room, or emergency department. Indications were abruption/previa (32.2%), congenital anemia (i.e., fetomaternal hemorrhage; 15.4%), umbilical cord accident (i.e., velamentous insertion; 15.0%), and bleeding/coagulopathy (12.8%). Fifty-eight percent of those with hemorrhage before birth did not have a hemoglobin value reported on the umbilical cord gas; thus, anemia was not recognized initially. None of the 149 ERBTs were administered using a blood warmer. The mortality rate of recipients was 35%. CONCLUSION: Based on our findings, we recommend including a hemoglobin value with every cord blood gas after emergency delivery to rapidly identify fetal anemia. We also discuss two potential improvements for future testing: 1) the use of a warming device for massive transfusion of neonates and 2) the use of low-titer group O cold-stored whole blood for massive hemorrhage in neonates.

The prevalence of an elevated F cell population in a maternal and gynaecology cohort.

OBJECTIVES: This study aimed to determine F cell prevalence in a cohort of maternal and gynaecology specimens using QuikQuant anti-HbF flow cytometry (FC) kit and to investigate if the presence of maternal F cells can lead to fetomaternal haemorrhage (FMH) overestimation. BACKGROUND: The gold standard to estimate FMH is the Kleihauer-Betke test (KBT). The KBT has proved to be insufficiently sensitive to detect low numbers of circulating fetal cells due to the presence of maternal F cells. At present, the prevalence of false positive KBT results due to raised maternal F cell population, defined as >5%, is poorly characterised. METHODS: A total of 120 specimens were tested for the presence of F cells and fetal cells by KBT and anti-HbF FC. The results calculated were compared to determine FMH overestimation. RESULTS: Of our cohort, 32% showed an elevated F cell population, of which 69% (27 of 39) were clinically significant according to KBT (>2 mL FMH). The mean FMH volumes by KBT and anti-HbF FC were 3·90 mL (0·20-35·40 mL) and 4·09 mL (0·20-9·70 mL), respectively. CONCLUSION: The study highlighted that an elevated F cell level could be found in the cohort tested, with an F cell level of >10% causing significant FMH overestimation by KBT.

Fetomaternal hemorrhage among pregnant women in Accra, Ghana.

OBJECTIVE: To determine the occurrence of and risk factors for fetomaternal hemorrhage (FMH) among pregnant women at Korle Bu Teaching Hospital in Accra, Ghana. METHODS: A prospective study of FMH among pregnant women without hemoglobinopathies in the second trimester attending prenatal care between October 2015 and May 2016 performed using the Kleihauer-Betke test. Volume of FMH was estimated; ABO and Rh blood groups of participants were determined. A data extraction form and structured questionnaire were used to collect demographic and clinical information, and data on risk factors. RESULTS: Of 151 participants, 32 (21.2%) had FMH. Almost 18% (n=27) had FMH at baseline (16-24 weeks), 10% (10/100) at 28-32 weeks, and 11.1% (11/99) at 34-37 weeks of pregnancy. Volume of FMH was less than 30 mL in 30 (19.9%) women, whereas it was greater than 30 mL in 2 (1.3%) women. No identifiable patient-specific factors were associated with occurrence of FMH. CONCLUSION: FMH is common among pregnant women in Ghana and can occur as early as 16 weeks, without identifiable risk factors. RhD negative women who may be pregnant with RhD positive fetuses should be screened early in pregnancy, not only at delivery, for occurrence of FMH.

Evaluation of Automatic Blood Analyzer as Screening Method in Fetomaternal Hemorrhage.

Screening of fetomaternal hemorrhage (FMH) is essential in management of fetomaternal antigen incompatibilities of blood. The objective in this study was to evaluate the ability of automatic blood analyzer (ABA) to screen FMH, also comparing this method with flow cytometry (FCM). The contents of fetal red blood cells and fetal hemoglobin were evaluated by FCM and ABA, respectively, using both blood samples of male adults laced with umbilical cord blood diluted at 1/10, 1/100, 1/1,000, and 1/10,000, or blood from puerperal women collected within 48 hours following delivery. FCM had better performance (area under curve, AUC = 0.8723) than ABA (AUC = 0.6569) in detecting fetal blood laced with blood from male adults. At a critical level of 0.5%, ABA indicated that 27.5% of puerperal women would have FMH while FCM did not detect FMH. Our results showed that ABA overestimates FMH and disagrees with FCM on indicating puerperal women with FMH. ABA is inadequate for being used to screen for or to measure FMH.

Enumeration of Fetal Red Blood Cells, Hemoglobin-Specific RBC Cells, and F Reticulocytes in Human Blood.

Recent advances in analytical cytometry have improved diagnostic tools for the study of erythropoiesis in anemic patients and resolution of differential diagnosis in diseases of the erythron. This article presents three applications of red blood cell (RBC) analysis-quantitation of fetal red cells, F-cell enumeration, and F-reticulocyte analysis-which improve diagnostic precision, sensitivity, and specificity, and provide better laboratory indicators of therapeutic efficacy in a variety of hematologic and obstetric disorders. Such advances also include the measurement and quantitation of RBC hemoglobins and their relative ribonucleic acid levels. These advances not only promise to improve diagnostic accuracy and laboratory precision over techniques such as the traditional manual reticulocyte counting method and the Kleihauer-Betke stain method for evaluating fetomaternal hemorrhage (FMH), but also serve as tools for newer assays of anemia diagnosis and improved clinical outcomes. In addition to the primary methods, supporting techniques for preparing spiked controls, automating data analysis, setting up a fetal hemoglobin acquisition protocol, and assaying reticulocytes using thiazole orange are also presented. © 2019 by John Wiley & Sons, Inc.

Prospective quantification of fetomaternal hemorrhage with dilation and evacuation procedures.

OBJECTIVE: To describe fetomaternal hemorrhage (FMH) during second-trimester dilation and evacuation (D&E) to evaluate if Rhesus-immune globulin (RhIG) 100 mcg (used in the United Kingdom) and 300 mcg (used in the United States) provide adequate prophylaxis. STUDY DESIGN: We conducted an exploratory prospective descriptive study of women undergoing D&E between 15 weeks 0 days and 23 weeks 6 days of gestation. Enrolled participants had Kleihauer-Betke testing on specimens obtained before and after D&E. We assessed the main outcome measures of FMH in mL suggesting need for more than 100 mcg and 300 mcg RhIG (FMH of 10 mL and 30 mL fetal whole blood, respectively) and association of postprocedure FMH with demographic characteristics and procedure-related variables. RESULTS: The 300 participants had a mean gestational age of 19 weeks 6 days±2 weeks 2 days. The median preprocedure FMH was 0 mL (range 0-50 mL) with 2 (0.67%) women exceeding 10 mL (19 mL and 50 mL). The median postprocedure FMH was 1 mL (range 0-60 mL). Almost all participants had postprocedure FMH <10 mL (n=295, 98.3%) and <30 mL (n=298, 99.3%). All participants under 18 weeks had FMH <10 mL. We found no demographic or procedure-related factors to be predictive of FMH quantity. CONCLUSIONS: FMH occurring with routine second-trimester D&E procedures is minimal. Adequate prophylaxis with RhIG 100 mcg and 300 mcg occurred in >98% of women and in all cases <18 weeks of gestation. This study is the first step to potentially reducing the dose and costs of RhIG administration with D&E. IMPLICATIONS: This study is a first step in quantifying fetomaternal hemorrhage with routine dilation and evacuation procedures; larger trials are needed, especially to understand why some women have recognizable hemorrhage preprocedure. If dosing requirements are too high with current guidelines, lower doses will result in resource and cost savings.

Does Maternal Body Mass Index Affect the Quantity of Circulating Fetal Cells Available to Use for Cell-Based Noninvasive Prenatal Test in High-Risk Pregnancies?

We present the first study that investigates the effect of maternal body mass index (BMI) on the quantity of circulating fetal cells available to use in cell-based noninvasive prenatal test (cbNIPT). cbNIPT has been proposed as a superior alternative to noninvasive prenatal test from cell-free fetal DNA. Kølvraa et al. [Prenat Diagn. 2016 Dec; 36(12): 1127-34] established that cbNIPT can be performed on as few as one fetal cell, and Vestergaard et al. [Prenat Diagn. 2017 Nov; 37(11): 1120-4] demonstrated that these fetal trophoblast cells could be used successfully in cbNIPT to detect chromosomal and sub-chromosomal abnormalities. This study on 91 pregnant women with high-risk pregnancies suggests that cbNIPT should not be hampered by an increased BMI because every pregnancy, irrespective of the BMI, has rendered fetal cells for downstream genetic analysis. The mean number of fetal cells per sample was 12.6, with a range of 1-43 cells in one sample. ANOVA showed that increasing maternal BMI tends to decrease the number of fetal cells, but not significantly.

Elimination of ABO Blood Group Incompatible Fetal Red Blood Cells in the Maternal Circulation: Relevance to the Diagnosis of Fetal-Maternal Hemorrhage.

BACKGROUND: Fetal-maternal hemorrhage (FMH) occurs when fetal red blood cells (RBC) pass into the maternal circulation as a result of obstetric- or trauma-related complications to pregnancy. Their detection in the maternal blood is commonly used as a diagnostic test. There is, however, a serious and general limitation to this test that is sometimes ignored. Fetal RBC carrying the father's antigens (most crucially, the ABO blood antigens) may be incompatible with the mother's plasma. They are expected to be eliminated by the maternal natural antibodies, thus, negative results may be false. OBJECTIVES: By simulating fetal-maternal ABO incompatibility, we studied the fate of fetal RBC in vitro. METHODS: Adult blood samples (n = 6) of O-blood group (type) were mixed with 1-5% cord blood or neonatal blood of A- or O-type, representing incompatible and compatible fetal RBC, respectively. The survival of fetal RBC was quantified after an overnight incubation. The supernatant was assayed for fetal hemoglobin (HbF) using the spectrophotometric alkaline-resistance benzidine assay, while the pellet was assayed for HbF/carbonic anhydrase (CA) expression in RBC by flow cytometry. The HbFhigh/CAlow phenotype characterizes fetal RBC. RESULTS: Both assays demonstrated disappearance of the fetal RBC due to lysis upon incubation in incompatible blood. CONCLUSIONS: A similar situation may also occur in vivo. Thus, under these conditions, negative results in the FMH test may be false, and lead to misdiagnosis.

Mild-to-moderate foeto-maternal haemorrhage in the third trimester and at term of pregnancy: quantitative determination and clinical-diagnostic evaluation.

BACKGROUND: Foeto-maternal haemorrhage (FMH), a gestational event that occurs before or during delivery, consists of a loss of foetal blood into the maternal circulation. FMH occurs more frequently during the third trimester or labour both in normal and complicated pregnancies. In the case of alloimmunisation, the maternal immunological response and the severity of the resulting foetal or neonatal disease depend on the amount of foetal blood that passes into the maternal circulation. The aim of this study was to determine FMH in the third trimester and at term of pregnancy and to evaluate the role of clinical and ultrasound markers in the prediction of FMH. MATERIALS AND METHODS: FMH was quantified by cytofluorimetric testing at 28 to 35 weeks of gestation in 223 women and at term in 465 women, all with risk factors. Foetal evaluation included foetal movement profile, middle cerebral artery peak velocity of systolic blood flow (MCA-PSV) and cardiotocographic monitoring. RESULTS: All women tested negative for FMH in the third trimester. Four patients (0.9%) tested positive at term, with estimated volumes of bleeding of 2.2, 8.1, 12.3 and 39.8 mL. Three FMH cases (75%) had a non-reassuring cardiotocography compared to 8.9% (42/461) of women without FMH (p=0.003) and two FMH cases reported a reduction in foetal movements reduction compared to four of those without FMH (p=0.001). Mean MCA-PSV was normal in both the groups with and without FMH (p=0.22). DISCUSSION: FMH is rare in pregnancy and at term. Cytofluorimetric testing is a specific method to detect mild-to-moderate FMH even when the MCA-PSV is not informative. Mild-to-moderate FMH is significantly associated with reduced foetal movements and non-reassuring cardiotocographic monitoring.

Clinical input of anti-D quantitation by continuous-flow analysis on autoanalyzer in the management of high-titer anti-D maternal alloimmunization.

BACKGROUND: In addition to titration by indirect antiglobulin test most widely used, anti-D quantitation by continuous-flow analysis (CFA) may be performed to assess severity of maternal immunization. Only five studies have reported its added value in the management of pregnancies complicated by anti-D immunization. STUDY DESIGN AND METHODS: A retrospective study of 74 severe anti-D-immunized pregnancies was conducted from January 1, 2013, to December 31, 2014, in the Trousseau Hospital in Paris (France). Concentration of maternal anti-D was measured by titration and by CFA two-stages method (2SM; total amount of anti-D) and one-stage method (1SM; high-affinity IgG1 anti-D). These biologic data were analyzed according to the severity of the hemolytic disease of the fetus and the newborn. RESULTS: The value of 5 IU anti-D/mL in maternal serum is validated as a threshold to trigger ultrasonographic and Doppler fetal close follow-up. A high 1SM/2SM ratio was associated with a higher risk of intrauterine transfusion (IUT). For pregnancies requiring IUT and without increasing titer, maternal 1SM anti-D concentration tends to correlate with the precocity of fetal anemia. In the "without-IUT" group 1SM and 2SM anti-D concentrations correlate significantly with cord bilirubin levels of the newborn at birth. CONCLUSION: Altogether our results underline the importance of anti-D quantitation by CFA to optimize the management of anti-D-alloimmunized pregnancies.

The presence of F cells with a fetal phenotype in adults with hemoglobinopathies limits the utility of flow cytometry for quantitation of fetomaternal hemorrhage.

BACKGROUND: Detection and quantitation of fetomaternal hemorrhage (FMH) can be difficult in patients with pre-existing elevations of HbF, such as those with hemoglobinopathies. The aim of this study was to evaluate the utility of dual-color flow cytometry with the Fetal Cell Count Kit (FCCK) in differentiating adult and fetal HbF in this population, as compared to flow cytometry (FC) using HbF alone. METHODS: Peripheral blood was obtained from normal adults and patients with hemoglobinopathies (ß-thalassemia and sickle cell disease), including a small number of pregnant females. Cord blood was used to spike some samples with 5% fetal cells. Analysis by single color (HbF) and dual-color (HbF and carbonic anhydrase) FC was performed on these samples. Fetal cells were defined as those with high HbF fluorescence on single-color FC, and those that were HbF + CA- using the FCCK. The quantity of fetal cells detected by each technique was compared. RESULTS: Forty-six adult patients were included. In non-pregnant adults with hemoglobinopathies, a population of red cells with a fetal cell phenotype were detected by both techniques. The dual-color method reported lower quantities of these cells. In nineteen samples spiked with cord blood the FCCK consistently underestimated the quantity of fetal cells. CONCLUSIONS: Patients with ß-thalassemia and sickle cell disease have a population of HbF-containing cells which are phenotypically similar to fetal cells. Even with dual-color flow cytometry (FCCK), the detection and quantification of FMH by flow cytometry in this population remains difficult. © 2017 International Clinical Cytometry Society.

A descriptive single-centre experience of the management and outcome of maternal alloantibodies in pregnancy.

BACKGROUND: Haemolytic disease of the fetus and newborn (HDFN) occurs when maternal IgG alloantibodies to fetal red blood cell antigens cross the placenta, causing haemolysis in the fetus and/or neonate. After delivery, the main concern is hyperbilirubinaemia, which can cause neurological damage. OBJECTIVES: To summarise our current management and outcome data to inform health-care professionals counselling women whose pregnancies are at risk of HDFN and to compare these data with relevant studies. METHODS: This is a retrospective descriptive study of all high-risk pregnancies at risk of HDFN at Guy's and St. Thomas' NHS Foundation Trust (GSTFT) Maternity Unit over a 7-year period. We defined high-risk pregnancies as those in whom anti-D, anti-c, anti-K or high (>32 or doubling strength) titres of all other antibodies were identified. RESULTS: A total of 130 pregnancies in 112 women were followed up. A single alloantibody was found in 93 pregnancies (71.5%) and multiple alloantibodies in 37 pregnancies (28.5%). Anti-D was most commonly encountered (n = 48, 36.9%), followed by anti-c (n = 31, 23.8%) and anti-E (n = 15, 11.5%). In 65 of 130 pregnancies (50%), antibody concentrations triggered scans to screen for fetal anaemia. Of 130 pregnancies, 6 (4.6%) required intrauterine transfusions, and 31 of 130 (26%) neonates required post-natal intervention. Overall, morbidity was 0.1% and mortality 0.002%. CONCLUSIONS: This study demonstrates that morbidity and mortality caused by HDFN is minimal. These results are reassuring for women at risk of HDFN as even severely affected cases are successfully managed in most instances. Further studies are needed to identify predictors of disease severity.

Blood Group Antigen Matching Influence on Gestational Outcomes (AMIGO) study.

BACKGROUND: Red blood cell (RBC) antigen matching policies to prevent alloimmunization in females of childbearing potential (FCP) vary between centers. To inform transfusion centers responsible for making decisions about matching policies for FCPs, the causal stimulus of the antibodies implicated in severe hemolytic disease of the fetus and newborn (HDFN) must be determined. STUDY DESIGN AND METHODS: We conducted a multinational retrospective study of women with offspring affected by severe HDFN requiring neonatal exchange transfusion and/or intrauterine transfusion. Mothers treated at centers that provide extended antigen-negative RBCs (MATCH, five centers) and those that do not (NoMATCH, nine centers) were compared. RESULTS: A total of 293 mothers had at least one affected pregnancy: 179 at MATCH centers and 114 at NoMATCH centers. Most alloimmunization (83%) was attributed to previous pregnancy: 3% to transfusion (two cases at MATCH, six at NoMATCH centers) and 14% undetermined (both antecedent transfusion and pregnancy). Only 50 mothers had received transfusions; 13 had HDFN due to anti-K at MATCH and four at NoMATCH centers. Most (12/13, 92%) of the anti-K HDFN cases at MATCH centers had K+ paternal antigen status. Mothers at the MATCH centers do not appear to be protected from HDFN due to K, C, c, and E antibodies, although the low number of FCPs who received transfusions precluded drawing firm conclusions. CONCLUSION: The causal stimulus of antibodies that cause HDFN is predominantly from previous pregnancy. Although extended RBC matching for FCPs may impart some protection from allosensitization, we were unable to show a positive effect, possibly because matching policies are not uniform and there was a small number of mothers who previously received transfusions.

Characterisation of maternal human leukocyte antigen class I antibodies in suspected foetal and neonatal alloimmune thrombocytopenia.

OBJECTIVES: To investigate the specificities and level of HLA class I antibodies in selected cases referred for suspected foetal and neonatal alloimmune thrombocytopenia (FNAIT). BACKGROUND: FNAIT occurs in 1 : 1-2000 live births, whereas maternal immunisation against human leukocyte antigen (HLA) class I is common. Whether HLA class I antibodies alone can cause FNAIT is debatable. MATERIAL AND METHODS: A total of 260 patient samples were referred between 2007 and 2012. Referrals with maternal HLA class I antibodies and no other cause for the neonatal thrombocytopenia were included for analysis (cases, n = 23). HPA-1a negative mothers were excluded. Control groups were screened positive mothers of healthy neonates (controls, n = 33) and female blood donors (blood donors, n = 19). LABScreen single antigen HLA class I beads was used for antibody analysis. Clinical records were reviewed for cases. RESULTS: All groups had broad antibody reactivity. Cases had more antibodies with high SFI levels compared with the controls (SFI>9999; medians 26, 6 and 0; P < 0·05) and higher overall median HLA-ABC and HLA-B SFI (P < 0·05). Many of the antibodies were reactive with rare alleles. When reviewing the clinical records, several of the cases had other contributing factors to the thrombocytopenia. There was no correlation between foetal platelet count and antibody levels. CONCLUSION: Mothers of thrombocytopenic neonates had higher levels of HLA class I antibodies compared with control groups of women with healthy children and female blood donors. However, clinical outcome and antibody response correlated poorly in the heterogeneous case group, indicating a multifactorial cause to the thrombocytopenia in the majority of cases.

Development of novel noninvasive prenatal testing protocol for whole autosomal recessive disease using picodroplet digital PCR.

We developed a protocol of noninvasive prenatal testing (NIPT), employing a higher-resolution picodroplet digital PCR, to detect genetic imbalance in maternal plasma DNA (mpDNA) caused by cell-free fetal DNA (cffDNA). In the present study, this approach was applied to four families with autosomal recessive (AR) congenital sensorineural hearing loss. First, a fraction of the fetal DNA in mpDNA was calculated. Then, we made artificial DNA mixtures (positive and negative controls) to simulate mpDNA containing the fraction of cffDNA with or without mutations. Next, a fraction of mutant cluster signals over the total signals was measured from mpDNA, positive controls, and negative controls. We determined whether fetal DNA carried any paternal or maternal mutations by calculating and comparing the sum of the log-likelihood of the study samples. Of the four families, we made a successful prediction of the complete fetal genotype in two cases where a distinct cluster was identified for each genotype and the fraction of cffDNA in mpDNA was at least 6.4%. Genotyping of only paternal mutation was possible in one of the other two families. This is the first NIPT protocol potentially applicable to any AR monogenic disease with various genotypes, including point mutations.

Red cell alloimmunization in RhD positive pregnant women and neonatal outcome.

The frequency of red blood cell (RBC) alloimmunization in RhD positive pregnant women is not known in our population. We planned to determine its frequency and correlation with neonatal outcome. We included 1000 RhD positive pregnant women: 500 had 'normal pregnancy' (Group I) and another 500 had 'high risk pregnancy' (Group II). ABO and extended Rh phenotyping were done by tube technique, antibody screening and identification by gel technique. For alloimmunized women, the paternal and neonatal ABO and extended Rh typing were done. Neonatal direct antiglobulin test (DAT) was also done and their clinical outcome observed. The frequency of RBC alloimmunization was 0.7% (7/1000) and all these women were from group II (p = 0.015). The alloantibodies were anti-E (85.7%), anti-c (71.4%), anti-Cw (14.3%) and anti-S (14.3%). Also, 6 women had history of transfusion (p < 0.01). Of the 7 neonates born to alloimmunized mothers, 4 (57.14%) had a positive DAT. The mean duration of phototherapy was higher in the DAT positive neonates (p < 0.01) and 2 (50%) required exchange transfusion. Thus, the frequency of alloimmunization was 0.7% in RhD positive pregnant women. High risk pregnancies and antenatal patients having a history of blood transfusion should be considered for regular antibody screening.