High sensitivity and specificity of a new functional flow cytometry assay for clinically significant heparin-induced thrombocytopenia antibodies

Abstract Background Capsule endoscopy (CE) plays a central role in the management of patients with suspected or known Crohn’s disease (CD). It is indicated for the diagnosis, classification, monitoring of the response to treatment, and prognostic prediction. In 2017, PillCam™ Crohn’s Capsule (PCC) was introduced. It has demonstrated greater accuracy in detecting and evaluating the extent of lesions in these patients. However, this new tool produces thousands of images, whose revision is time-consuming and prone to errors, since lesions can be restricted to a small number of images. In the last decade, several Artificial Intelligence (AI) algorithms were developed, and demonstrated potential to mitigate some of the drawbacks of CE. Among AI tools, Convolutional Neural Networks (CNN) display the best performance for imagery analysis. This study aims to develop an AI algorithm based on an CNN for the automatic detection of ulcers and erosions of the small intestine and colon in PCC images. Methods A total of 8 085 PCC images were extracted from a single tertiary centre between 2017–2020. This pool of images was constituted by 2 855 images depicting ulcers, 1 975 erosions; the remaining with normal enteric and colonic mucosa. For the automatic identification of these findings, this pool of images was split into training and validation datasets. A CNN model with transfer learning using tensorflow and keras tools was constructed. The performance of the network was subsequently assessed in an independent test set. Results After optimizing the different layers of the CNN, our model was able to detect and distinguish small intestinal or colonic erosions or ulcers with a sensitivity and specificity of 90.0% and 96.0%, respectively. The precision and accuracy of this model were 97.1% and 92.4%, respectively (Figure 1). Particularly, the CNN detected ulcers with a sensitivity of 83% and specificity of 98%, and erosions with sensitivity and specificity of 91% and 93%, respectively. Conclusion Our group developed, for the first time, a CNN capable of automatically detecting ulcers and erosions of the small intestine and colon in PCC images with high sensitivity and specificity. These findings are extremely important since they pave the way for the development of systems for the automatic detection of clinically significant lesions, optimizing diagnostic performance and efficiency of monitoring CD activity.

Download Full-text

High sensitivity and specificity screening for clinically significant intimate partner violence.

Journal of Family Psychology ◽

10.1037/fam0000781 ◽

2020 ◽

Author(s):

Richard E. Heyman ◽

Katherine J. W. Baucom ◽

Shu Xu ◽

Amy M. Smith Slep ◽

Jeffery D. Snarr ◽

...

Keyword(s):

Intimate Partner Violence ◽

Sensitivity And Specificity ◽

Partner Violence ◽

High Sensitivity ◽

Intimate Partner ◽

Clinically Significant

Download Full-text

High-Sensitivity Flow Cytometry Testing for Paroxysmal Nocturnal Hemoglobinuria in Children with Cytopenia: A Single Center Study

Blood ◽

10.1182/blood.v118.21.2398.2398 ◽

2011 ◽

Vol 118 (21) ◽

pp. 2398-2398

Author(s):

Choladda V. Curry ◽

M. Tarek Elghetany ◽

Andrea M. Sheehan ◽

Alison A. Bertuch ◽

Ghadir S. Sasa

Keyword(s):

Bone Marrow ◽

Flow Cytometry ◽

Aplastic Anemia ◽

Immunosuppressive Therapy ◽

Hemolytic Anemia ◽

Paroxysmal Nocturnal Hemoglobinuria ◽

High Sensitivity ◽

Deficiency Anemia ◽

Unknown Etiology ◽

Flow Cytometry Assay

Abstract Abstract 2398 Background: Paroxysmal nocturnal hemoglobinuria (PNH) is a rare acquired hematopoietic stem cell disorder characterized by expansion of cells with complete or partial loss of glycosyl phosphatidyl-inositol-anchored proteins. PNH usually presents with one or more of three clinical manifestations: intravascular hemolysis, thrombosis, or acquired bone marrow failure [aplastic anemia (AA) or myelodysplastic syndrome (MDS)]. Flow cytometry has become the gold standard for the diagnosis of PNH, particularly with the recent publication of guidelines for the diagnosis and monitoring of PNH and related disorders in 2010. PNH occurs rarely in children, and, consequently, the published literature regarding PNH in this pediatric population consists only of small case series, making it difficult to extrapolate the frequency of which PNH clones are identified. Moreover, no studies are available on the incidence of PNH clones in children with MDS and acquired aplastic anemia (AAA). We, therefore, sought to determine how frequently a high sensitivity FLAER-based assay, with a sensitivity of 0.01%, would detect PNH clones in children with cytopenias. Method and Results: The study period was from December 2010 to July 2011. PNH testing was performed using a high sensitivity FLAER based assay according to published guidelines using the combination of FLAER/CD64/CD15/CD33/CD24/CD14/CD45 for WBC testing and CD235a/CD59 for RBC testing. There were 31 peripheral blood samples from 29 patients (17 males/12 females) ranging in age from 4 months to 17 years (median, 10 years). All patients were tested for PNH because of cytopenia [pancytopenia (n = 14) and uni- or bicytopenia (n = 15)]. Patients had a mean Hgb of 10.7 gm/dL, mean ANC of 2.66 X103/uL and mean platelet of 115 X103/uL. Review of medical charts revealed the following clinical diagnoses: classic PNH - episodic hemolytic anemia with persistent thrombocytopenia (1), severe AA (SAA, 8), SAA with myelofibrosis (1), MDS (1), Fanconi anemia (1), chronic thrombocytopenia (2), refractory iron deficiency anemia (1), bone marrow suppression likely due to virus/medication (1), parvovirus infection (1), Copper deficiency (1), systemic lupus erythematosus (SLE, 1), and cytopenia of unknown etiology (10). Of note, all patients with AAA had SAA. PNH clones were identified in 6 out of 29 patients (20%): minor clones (<1% PNH population) in 3 patients: average clone sizes 0.12% [range 0.02–0.25] granulocytes (G), 0.51% [0.20–0.99] monocytes (M), and 0.08% [0.04–0.14] red blood cells (RBCs), and major clones (>1% PNH population) in 3 patients: average clone sizes 31.11% [3.98–67.58] G, 31.98% [6.15–71.1] M, and 14.76% [1.19–38.03] RBC, respectively, with ages ranging from 4 to 17 years. Patients who were identified to have minor PNH clones all presented with pancytopenia. Two were diagnosed with SAA; the cause of pancytopenia in the third patient is currently under investigation. None of patients with minor PNH clones had evidence of hemolysis or thrombosis. The three patients with major PNH clones had the following: Classic PNH with hemolytic anemia (1), SAA with PNH clones detected at the time of SAA diagnosis (1), and SAA with PNH clones detected 20 months after immunosuppressive therapy (1). The latter two patients did not have evidence of hemolysis or thrombosis. Of the 10 patients with a diagnosis of SAA or MDS, PNH clones were identified in 4 (40%) patients (2 with minor clones, 2 with major clones). Conclusions: This is the first study to describe the utility of using a standardized high-sensitivity FLAER-based flow cytometry assay to identify PNH clones in children. This is also the first study describing the prevalence of PNH clones in children with MDS and AAA. The identification of a PNH population in 40% of the MDS and AAA cases emphasizes the need for PNH testing in all children with these disorders using a high-sensitivity FLAER based flow cytometry assay. A low sensitivity assay would have missed 2 patients with minor PNH clones. This finding may be of significance considering SAA or MDS patients with PNH clones are more likely to respond to immunosuppressive therapy. Further studies are needed to investigate the prevalence of PNH clones in this setting and its impact on disease manifestations, course, and outcomes in children. Disclosures: No relevant conflicts of interest to declare.

Download Full-text

The sensitivity and specificity of the Modified Conflict Tactics Scale for detecting clinically significant elder abuse

International Psychogeriatrics ◽

10.1017/s1041610209009387 ◽

2009 ◽

Vol 21 (4) ◽

pp. 774-778 ◽

Cited By ~ 20

Author(s):

Claudia Cooper ◽

Kate Maxmin ◽

Amber Selwood ◽

Martin Blanchard ◽

Gill Livingston

Keyword(s):

Sensitivity And Specificity ◽

Clinical Judgment ◽

Scoring System ◽

Elder Abuse ◽

High Sensitivity ◽

Family Carers ◽

Conflict Tactics Scale ◽

People With Dementia ◽

Clinically Significant ◽

Conflict Tactics

ABSTRACTBackground: A third of family carers of people with dementia describe acting abusively in research studies, but far fewer cases of abuse are currently detected in clinical situations. This discrepancy may be explained by inadequate detection by health professionals, or disagreement regarding what constitutes elder abuse. This study was undertaken to determine the sensitivity and specificity of the revised Modified Conflict Tactics Scale (MCTS) for detecting clinically significant abuse.Methods: We interviewed 220 family carers of people consecutively referred to psychiatric services with dementia in Essex and London (U.K.), using the MCTS to measure abuse. We defined abuse cases using (1) the MCTS conventional scoring system; (2) the Pillemer criteria; and (3) clinical judgment of an expert panel.Results: Our panel judged that 15 (6.8%) of carers reported potentially clinical concerning abusive behavior; but 47 (21%) were cases according to the Pillemer criteria and 74 (34%) using the MCTS conventional scoring system. We developed a weighted MCTS scoring system, with high sensitivity and specificity for detecting clinically concerning abuse.Conclusions: The MCTS could be used routinely in clinical practice with carers of people with dementia to detect clinically concerning cases of abuse, many of which are currently being missed.

Download Full-text

PNH Clone Size By Flow Cytometry and Its Correlation with PIG Gene Mutation

Blood ◽

10.1182/blood-2018-99-115621 ◽

2018 ◽

Vol 132 (Supplement 1) ◽

pp. 4889-4889

Author(s):

Dajeong Jeong ◽

Hee Sue Park ◽

Seongmin Choi ◽

Heewon Seo ◽

Sung-Min Kim ◽

...

Keyword(s):

Flow Cytometry ◽

Gene Mutation ◽

Conflicts Of Interest ◽

Gene Mutations ◽

High Sensitivity ◽

Clone Size ◽

Definitive Evidence ◽

Flow Cytometric Measurement ◽

Clinically Significant ◽

Pnh Clone

Abstract Background: Routine diagnosis of paroxysmal nocturnal hemoglobinuria (PNH) is based on flow cytometric measurement of PNH clone in RBC and granulocyte with sensitivity of 1%, and recent introduction of high-sensitive flow cytometry (FCM) makes it possible to detect 0.01% clone. However, the clinical significance of small PNH clone has not been elucidated, so even the necessity to treat or not is controversial. We investigated whether quantitative results of PNH clone size measured by FCM correlate with mutant burden of PIG gene. Methods: A total of 44 specimens from 20 patients whose PNH clone size was >1% in either RBC or granulocyte by routine FCM were enrolled (classical PNH n=10, AA/PNH n=8, MDS/PNH n=2). To detect small quantity of cells with PIG gene mutation, we performed ultra-deep sequencing (average depth 3000X) for PIGA, PIGM, PIGT, and PIGX genes on these 44 consecutive specimens. Results: Sixteen patients (80.0%) were found to harbor PIG gene mutations: 15 patients had PIGA mutation and 1 patient had PIGM mutation. Granulocyte PNH clone size and variant allele frequency (VAF) of PIG gene mutation showed higher correlation (Spearman's r=0.73, p=0.0002) than that of RBC (Spearman's r=0.61, p=0.0073). All patients harboring PIG gene mutation showed more than 10% PNH clone by FCM. In contrast, 4 patients who did not have PIG gene mutation showed less than 10% PNH clone by FCM. Conclusion: Considering the mechanism of PNH development, the presence of PIG gene mutation is a definitive evidence supporting PNH diagnosis. PNH granulocyte clone size more than 10% by FCM seems to be clinically significant in relevance to PIG gene mutation. We suggest that neither high sensitivity FCM nor the detection of PIG gene mutation is a requisite for routine clinical diagnosis of PNH. Disclosures No relevant conflicts of interest to declare.

Download Full-text

Identification and Clinical Significance of Type II Granulocytes Among Patients with Paroxysmal Nocturnal Hemoglobinuria (PNH) Identified Using Multiparameter High-Sensitivity Flow Cytometry.

Blood ◽

10.1182/blood.v114.22.3015.3015 ◽

2009 ◽

Vol 114 (22) ◽

pp. 3015-3015 ◽

Cited By ~ 4

Author(s):

Mayur K Movalia ◽

Andrea Illingworth

Keyword(s):

Flow Cytometry ◽

High Sensitivity ◽

Type I ◽

Type Ii ◽

Flow Cytometry Assay ◽

Clone Size ◽

Type Iii ◽

Gpi Proteins ◽

Terminal Complement ◽

Pnh Clone

Abstract Abstract 3015 Poster Board II-991 PNH is a hematopoietic stem cell disorder in which unregulated activation of terminal complement leads to impaired quality of life and significant ischemic morbidities with shortened lifespan. Life-threatening thromboembolism (TE) is the most feared complication of PNH, accounting for 45% of patient deaths. Thrombosis has been observed in PNH patients regardless of the level of hemolysis. Additionally, platelet activation with subsequent consumption and thrombocytopenia are observed more often in PNH patients at risk for thrombosis. Current laboratory PNH diagnostic methods rely on flow cytometry to characterize PNH clones. PNH granulocytes (Gran) are typically detected using antibodies to a variety of GPI-linked markers including CD55, CD59, CD16, CD24, and CD66b. Recently, FLAER, a fluorescent proaerolysin variant that binds directly to the GPI anchor, has been used to identify and quantify GPI-deficient WBCs at a very high level of sensitivity. Although these markers are well established to detect granulocytes with normal expression of GPI proteins (Type I cells) and complete loss of GPI proteins (Type III cells), less is known about their ability to detect granulocytes with partial loss of GPI proteins (Type II cells). The ability to detect both PNH Type II RBCs and WBCs would provide clinically important information since quantitation of only PNH RBC clones can be confounded by transfusion or hemolysis. We evaluated 2,921 consecutive patient peripheral blood samples submitted for PNH diagnostic testing with a high-sensitivity flow cytometry assay for granulocytes that includes the fluorescent proaerolysin variant (FLAER) with confirmatory lineage-specific antibodies to GPI-linked antigens to distinguish Type I, II and III Gran clones. In addition, standard CD235/CD59 analysis was performed on the RBCs and evaluation with FLAER, CD14 and lineage-specific antibodies was performed on the monocytes. 216 patient samples (7.4%) had a detectable PNH gran clone (≥ 0.01% PNH Type III granulocytes and an absolute count of at least 50 cells). Clinical information was available for 162 of these patients (Table I). Of these samples, nineteen (8.8 %) patients demonstrated a distinct Type II Gran population, ranging in size from 1.2 – 65% (median clone size = 7%). In 4/19 patients, this Type II Gran population represented >50% of the total (Type II + Type III) PNH cells. In 10/19 patients (53%), a type II monocyte population was identified. Evaluation of the granulocyte markers (Table II) showed that the type II gran population was detectable in all cases by FLAER and in decreasing percentage by CD66b (88%), CD55 (50%), CD24 (47%) and CD16 (0%). Patients with Type II Gran clones had a significantly larger median total Gran PNH clone size than those without Type II Gran clones (87% vs. 11%; p= 0.0003), as well as larger median Type II and Type III RBC clones, likely a reflection of the ability to detect type II gran PNH clones with overall larger PNH clone sizes. Patients with Type II Gran clones showed significantly lower median platelet (plt) counts (54 ×109/L) than patients without Type II Gran clones (116×109/L; p< 0.01). Patients with Type II Gran clones had similar peripheral WBC, peripheral RBC, absolute neutrophil count, and hemoglobin (Hgb) compared to patients without Type II Gran clones, suggesting that differences in platelet counts are likely not due to differences in underlying marrow blood cell production. Type II PNH cells are an important component of the PNH diagnostic evaluation and both RBC and Gran Type II clones should be enumerated. In a large population of patients tested for the presence of PNH clones using a high sensitivity flow cytometry assay, a significant proportion of patients were identified with Type II PNH Gran clones. This study identified FLAER as the best reagent to identify type II Gran PNH clones and showed CD16 was least useful. This study also identified a clinical association between the presence of significant Type II clones and thrombocytopenia, potentially indicative of terminal complement-mediated platelet consumption. These findings are consistent with an increased risk of thrombosis in patients with significant Type II PNH clones. Disclosures: No relevant conflicts of interest to declare.

Download Full-text

High sensitivity 8-color flow cytometry assay for paroxysmal nocturnal hemoglobinuria granulocyte and monocyte detections

Biomedical Reports ◽

10.3892/br.2018.1047 ◽

2018 ◽

Author(s):

Ray Chun-Fai Chan ◽

Richard Leung ◽

Albert Posadas ◽

Thomas Lorey ◽

Allison Shaw

Keyword(s):

Flow Cytometry ◽

Paroxysmal Nocturnal Hemoglobinuria ◽

High Sensitivity ◽

Flow Cytometry Assay ◽

Color Flow

Download Full-text

A Multiplex Flow Cytometry Bead Assay For Simultaneous Detection Of IgG, IgA, IgM Heparin-Dependent Antibodies

Blood ◽

10.1182/blood.v122.21.3551.3551 ◽

2013 ◽

Vol 122 (21) ◽

pp. 3551-3551

Author(s):

Mia Sullivan ◽

Scott Grady ◽

Janice G McFarland ◽

Brian R Curtis

Keyword(s):

Flow Cytometry ◽

Sensitivity And Specificity ◽

Simultaneous Detection ◽

High Sensitivity ◽

Thrombin Inhibitors ◽

High Dose ◽

Good Sensitivity ◽

Antibody Isotype ◽

Igm Antibodies ◽

Dose Heparin

Abstract Heparin-Induced Thrombocytopenia (HIT) is an adverse reaction to heparin, in which affected patients produce platelet activating antibodies specific for complexes of platelet-factor 4 (PF4) and heparin (HPF4) resulting in a prothrombotic condition that in severe cases can be life-threatening. HIT is not rare, since as many as 600K people/year are affected. It is therefore important that a quick and accurate diagnosis of HIT be made, which requires both compatible clinical findings and serologic evidence of a heparin dependent antibody. First line testing consists of immunologic assays that detect IgG, IgA and IgM antibodies targeting PF4 in a complex with heparin or another polyanion. The most widely used assay is the PF4 ELISA, in which HPF4 bound to the wells of a microtiter plate is the target antigen. The ELISA has high sensitivity (near 100%) for antibody detection but a low positive predictive value for HIT diagnosis. Therefore, samples giving positive ELISA results are often retested using a functional assay like the serotonin release assay (SRA) that is more specific for HIT diagnosis. However, the SRA is difficult to perform and interpret and is not widely available. Further, the ELISA is time-consuming, costly, and most versions cannot distinguish IgG, IgA, IgM (IgG/A/M) antibodies. Knowledge of the antibody isotype can be useful when ELISA and SRA results are discrepant. For instance, strong ELISA results due to clinically insignificant IgM antibodies would not be detected by SRA, and additional testing to confirm IgM antibody isotype can lead to a delay in HIT diagnosis, and in some cases continuation of high cost treatments with direct thrombin inhibitors (DTI). Moreover, there is evidence that some IgA and IgM antibodies may cause HIT. Despite the high sensitivity of the ELISA and high specificity of the SRA, the shortcomings mentioned for these tests highlight the need for improved methods to support a diagnosis of HIT. We report development of a multiplex flow cytometry bead assay (MFBA) for simultaneous detection of IgG/A/M heparin-dependent antibodies that allows higher throughput and has sensitivity and specificity comparable to the PF4 ELISA. For the MFBA, polystyrene beads, to which chemically depolymerized heparin is covalently linked are incubated with purified platelet PF4 to produce HPF4-bead targets. Lyophilized HPF4 beads are stable for at least 3 months. HPF4-beads are incubated with serum, washed, and incubated with anti-human -IgG, -IgA, and –IgM, each with different fluorescent labels for detection of HIT antibodies by flow cytometry. A total of 164 patient samples could be tested by a single technologist in 8 hours. Serum samples from 300 patients with suspected HIT and previously tested by SRA were tested by both MFBA and an in-house PF4 ELISA including high dose heparin (100 U/L) confirmation step. Studies were approved by institutional IRB. MFBA results showed good sensitivity (87% IgG, 63% IgA, 36% IgM) for IgG, good specificity (96% IgG, 91% IgA, 93% IgM) for all isotypes when compared to the ELISA, and good sensitivity/specificity (93%/91%) compared to SRA. The area under the curve (AUC) calculated for the receiver operating characteristic graphs compared to the SRA were excellent (0.9-1.0) for both MFBA (AUC = 0.95) and PF4 ELISA (AUC = 0.98). All samples positive by MFBA inhibited with high dose heparin suggesting that this extra step is not required, and 3 samples that tested IgG positive in both MFBA and SRA gave indeterminate results (not inhibited with high dose heparin) in ELISA. Overall, these data show that the MFBA has accuracy, sensitivity, and specificity for detection of IgG antibodies comparable to the PF4 ELISA and the “gold standard” SRA with the additional advantage of simultaneously distinguishing IgG/A/M antibody isotypes, which will lead to reduced confirmation testing. Sensitivity of MFBA for detection of IgA and IgM antibodies was significantly lower compared to ELISA, but this reflects the assay design intended to increase specificity of the MFBA for these two isotypes. In addition to its multiplex capabilities, the MFBA allows for higher sample throughput and shortened turn-around times that will contribute to overall improved patient care. Disclosures: No relevant conflicts of interest to declare.

Download Full-text