Artificial intelligence-based predictions in neovascular age-related macular degeneration.

PURPOSE OF REVIEW: Predicting treatment response and optimizing treatment regimen in patients with neovascular age-related macular degeneration (nAMD) remains challenging. Artificial intelligence-based tools have the potential to increase confidence in clinical development of new therapeutics, facilitate individual prognostic predictions, and ultimately inform treatment decisions in clinical practice. RECENT FINDINGS: To date, most advances in applying artificial intelligence to nAMD have focused on facilitating image analysis, particularly for automated segmentation, extraction, and quantification of imaging-based features from optical coherence tomography (OCT) images. No studies in our literature search evaluated whether artificial intelligence could predict the treatment regimen required for an optimal visual response for an individual patient. Challenges identified for developing artificial intelligence-based models for nAMD include the limited number of large datasets with high-quality OCT data, limiting the patient populations included in model development; lack of counterfactual data to inform how individual patients may have fared with an alternative treatment strategy; and absence of OCT data standards, impairing the development of models usable across devices. SUMMARY: Artificial intelligence has the potential to enable powerful prognostic tools for a complex nAMD treatment landscape; however, additional work remains before these tools are applicable to informing treatment decisions for nAMD in clinical practice.

A Phase 2 Randomized, Double-Blind, Placebo-Controlled Trial of MHAA4549A, a Monoclonal Antibody, plus Oseltamivir in Patients Hospitalized with Severe Influenza A Virus Infection.

For patients hospitalized with severe influenza A virus infection, morbidity and mortality remain high. MHAA4549A, a human monoclonal antibody targeting the influenza A virus hemagglutinin stalk, has demonstrated pharmacological activity in animal studies and in a human influenza A challenge study. We evaluated the safety and efficacy of MHAA4549A plus oseltamivir against influenza A virus infection in hospitalized patients. The CRANE trial was a phase 2b randomized, double-blind, placebo-controlled study of single intravenous (i.v.) doses of placebo, 3,600 mg MHAA4549A, or 8,400 mg MHAA4549A each combined with oral oseltamivir (+OTV) in patients hospitalized with severe influenza A virus infection. Patients, enrolled across 68 clinical sites in 18 countries, were randomized 1:1:1. The primary outcome was the median time to normalization of respiratory function, defined as the time to removal of supplemental oxygen support to maintain a stable oxygen saturation (SpO2) of ≥95%. Safety, pharmacokinetics, and effects on influenza viral load were also assessed. One hundred sixty-six patients were randomized and analyzed during a preplanned interim analysis. Compared to placebo+OTV, MHAA4549A+OTV did not significantly reduce the time to normalization of respiratory function (placebo+OTV, 4.28 days; 3,600 mg MHAA4549A+OTV, 2.78 days; 8,400 mg MHAA4549A+OTV, 2.65 days), nor did it improve other secondary clinical outcomes. Adverse event frequency was balanced across cohorts. MHAA4549A+OTV did not further reduce viral load versus placebo+OTV. In hospitalized patients with influenza A virus infection, MHAA4549A did not improve clinical outcomes over OTV alone. Variability in patient removal from oxygen supplementation limited the utility of the primary endpoint. Validated endpoints are needed to assess novel treatments for severe influenza A virus infection. (This study has been registered at ClinicalTrials.gov under registration no. NCT02293863.).

Phase 2 Randomized Trial of the Safety and Efficacy of MHAA4549A, a Broadly Neutralizing Monoclonal Antibody, in a Human Influenza A Virus Challenge Model.

MHAA4549A, a human monoclonal antibody targeting the hemagglutinin stalk region of influenza A virus (IAV), is being developed as a therapeutic for patients hospitalized with severe IAV infection. The safety and efficacy of MHAA4549A were assessed in a randomized, double-blind, placebo-controlled, dose-ranging study in a human IAV challenge model. One hundred healthy volunteers were inoculated with A/Wisconsin/67/2005 (H3N2) IAV and, 24 to 36 h later, administered a single intravenous dose of either placebo, MHAA4549A (400, 1,200, or 3,600 mg), or a standard oral dose of oseltamivir. Subjects were assessed for safety, pharmacokinetics (PK), and immunogenicity. The intent-to-treat-infected (ITTI) population was assessed for changes in viral load, influenza symptoms, and inflammatory biomarkers. MHAA4549A was well tolerated in all IAV challenge subjects. The 3,600-mg dose of MHAA4549A significantly reduced the viral burden relative to that of the placebo as determined by the area under the curve (AUC) of nasopharyngeal virus infection, quantified using quantitative PCR (98%) and 50% tissue culture infective dose (TCID50) (100%) assays. Peak viral load, duration of viral shedding, influenza symptom scores, mucus weight, and inflammatory biomarkers were also reduced. Serum PK was linear with a half-life of ∼23 days. No MHAA4549A-treated subjects developed anti-drug antibodies. In conclusion, MHAA4549A was well tolerated and demonstrated statistically significant and substantial antiviral activity in an IAV challenge model. (This study has been registered at ClinicalTrials.gov under identifier NCT01980966.).

Human Monoclonal Antibody 81.39a Effectively Neutralizes Emerging Influenza A Viruses of Group 1 and 2 Hemagglutinins.

The pandemic threat posed by emerging zoonotic influenza A viruses necessitates development of antiviral agents effective against various antigenic subtypes. Human monoclonal antibody (hMAb) targeting the hemagglutinin (HA) stalk offers a promising approach to control influenza virus infections. Here, we investigated the ability of the hMAb 81.39a to inhibit in vitro replication of human and zoonotic viruses, representing 16 HA subtypes. The majority of viruses were effectively neutralized by 81.39a at a 50% effective concentration (EC50) of <0.01 to 4.9 µg/ml. Among group 2 HA viruses tested, a single A(H7N9) virus was not neutralized at 50 µg/ml; it contained HA2-Asp19Gly, an amino acid position previously associated with resistance to neutralization by the group 2 HA-neutralizing MAb CR8020. Notably, among group 1 HA viruses, H11-H13 and H16 subtypes were not neutralized at 50 µg/ml; they shared the substitution HA2-Asp19Asn/Ala. Conversely, H9 viruses harboring HA2-Asp19Ala were fully susceptible to neutralization. Therefore, amino acid variance at HA2-Asp19 has subtype-specific adverse effects on in vitro neutralization. Mice given a single injection (15 or 45 mg/kg of body weight) at 24 or 48 h after infection with recently emerged A(H5N2), A(H5N8), A(H6N1), or A(H7N9) viruses were protected from mortality and showed drastically reduced lung viral titers. Furthermore, 81.39a protected mice infected with A(H7N9) harboring HA2-Asp19Gly, although the antiviral effect was lessened. A(H1N1)pdm09-infected ferrets receiving a single dose (25 mg/kg) had reduced viral titers and showed less lung tissue injury, despite 24- to 72-h-delayed treatment. Taken together, this study provides experimental evidence for the therapeutic potential of 81.39a against diverse influenza A viruses. IMPORTANCE: Zoonotic influenza viruses, such as A(H5N1) and A(H7N9) subtypes, have caused severe disease and deaths in humans, raising public health concerns. Development of novel anti-influenza therapeutics with a broad spectrum of activity against various subtypes is necessary to mitigate disease severity. Here, we demonstrate that the hemagglutinin (HA) stalk-targeting human monoclonal antibody 81.39a effectively neutralized the majority of influenza A viruses tested, representing 16 HA subtypes. Furthermore, delayed treatment with 81.39a significantly suppressed virus replication in the lungs, prevented dramatic body weight loss, and increased survival rates of mice infected with A(H5Nx), A(H6N1), or A(H7N9) viruses. When tested in ferrets, delayed 81.39a treatment reduced viral titers, particularly in the lower respiratory tract, and substantially alleviated disease symptoms associated with severe A(H1N1)pdm09 influenza. Collectively, our data demonstrated the effectiveness of 81.39a against both seasonal and emerging influenza A viruses.

Deep Learning to Predict Geographic Atrophy Area and Growth Rate from Multimodal Imaging.

OBJECTIVE: To develop deep learning models for annualized geographic atrophy (GA) growth rate prediction using fundus autofluorescence (FAF) images and spectral-domain OCT volumes from baseline visits, which can be used for prognostic covariate adjustment to increase power of clinical trials. DESIGN: This retrospective analysis estimated GA growth rate as the slope of a linear fit on all available measurements of lesion area over a 2-year period. Three multitask deep learning models-FAF-only, OCT-only, and multimodal (FAF and OCT)-were developed to predict concurrent GA area and annualized growth rate. PARTICIPANTS: Patients were from prospective and observational lampalizumab clinical trials. METHODS: The 3 models were trained on the development data set, tested on the holdout set, and further evaluated on the independent test sets. Baseline FAF images and OCT volumes from study eyes of patients with bilateral GA (NCT02247479; NCT02247531; and NCT02479386) were split into development (1279 patients/eyes) and holdout (443 patients/eyes) sets. Baseline FAF images from study eyes of NCT01229215 (106 patients/eyes) and NCT02399072 (169 patients/eyes) were used as independent test sets. MAIN OUTCOME MEASURES: Model performance was evaluated using squared Pearson correlation coefficient (r2) between observed and predicted lesion areas/growth rates. Confidence intervals were calculated by bootstrap resampling (B = 10 000). RESULTS: On the holdout data set, r2 (95% confidence interval) of the FAF-only, OCT-only, and multimodal models for GA lesion area prediction was 0.96 (0.95-0.97), 0.91 (0.87-0.95), and 0.94 (0.92-0.96), respectively, and for GA growth rate prediction was 0.48 (0.41-0.55), 0.36 (0.29-0.43), and 0.47 (0.40-0.54), respectively. On the 2 independent test sets, r2 of the FAF-only model for GA lesion area was 0.98 (0.97-0.99) and 0.95 (0.93-0.96), and for GA growth rate was 0.65 (0.52-0.75) and 0.47 (0.34-0.60). CONCLUSIONS: We show the feasibility of using baseline FAF images and OCT volumes to predict individual GA area and growth rates using a multitask deep learning approach. The deep learning-based growth rate predictions could be used for covariate adjustment to increase power of clinical trials. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found after the references.

Geographic Atrophy Segmentation Using Multimodal Deep Learning.

Purpose: To examine deep learning (DL)-based methods for accurate segmentation of geographic atrophy (GA) lesions using fundus autofluorescence (FAF) and near-infrared (NIR) images. Methods: This retrospective analysis utilized imaging data from study eyes of patients enrolled in Proxima A and B (NCT02479386; NCT02399072) natural history studies of GA. Two multimodal DL networks (UNet and YNet) were used to automatically segment GA lesions on FAF; segmentation accuracy was compared with annotations by experienced graders. The training data set comprised 940 image pairs (FAF and NIR) from 183 patients in Proxima B; the test data set comprised 497 image pairs from 154 patients in Proxima A. Dice coefficient scores, Bland-Altman plots, and Pearson correlation coefficient (r) were used to assess performance. Results: On the test set, Dice scores for the DL network to grader comparison ranged from 0.89 to 0.92 for screening visit; Dice score between graders was 0.94. GA lesion area correlations (r) for YNet versus grader, UNet versus grader, and between graders were 0.981, 0.959, and 0.995, respectively. Longitudinal GA lesion area enlargement correlations (r) for screening to 12 months (n = 53) were lower (0.741, 0.622, and 0.890, respectively) compared with the cross-sectional results at screening. Longitudinal correlations (r) from screening to 6 months (n = 77) were even lower (0.294, 0.248, and 0.686, respectively). Conclusions: Multimodal DL networks to segment GA lesions can produce accurate results comparable with expert graders. Translational Relevance: DL-based tools may support efficient and individualized assessment of patients with GA in clinical research and practice.

A Phase 2 Randomized, Double-Blind, Placebo-Controlled Trial of the Monoclonal Antibody MHAA4549A in Patients With Acute Uncomplicated Influenza A Infection.

BACKGROUND: MHAA4549A, a human monoclonal antibody targeting the influenza A hemagglutinin stalk, neutralizes influenza A virus in animal and human volunteer challenge studies. We investigated the safety and tolerability, efficacy, and pharmacokinetics of MHAA4549A in outpatients with acute, uncomplicated influenza A infection. METHODS: This was a phase 2, randomized, double-blind, placebo-controlled trial of single intravenous (IV) doses of 3600 mg or 8400 mg of MHAA4549A or IV placebo in adult outpatients testing positive for influenza A. Patients were enrolled across 35 sites in 6 countries. Randomization and dosing occurred within ≤72 hours of symptom onset; the study duration was 14 weeks. The primary end point was the nature and frequency of adverse events (AEs). Secondary end points included median time to alleviation of all influenza symptoms, effects on nasopharyngeal viral load and duration of viral shedding, and MHAA4549A serum pharmacokinetics. RESULTS: Of 125 randomized patients, 124 received study treatment, with 99 confirmed positive for influenza A by central testing. The frequency of AEs between the MHAA4549A and placebo groups was similar; nausea was most common (8 patients; 6.5%). MHAA4549A serum exposure was confirmed in all MHAA4549A-treated patients and was dose-proportional. No hospitalizations or deaths occurred. Between the MHAA4549A and placebo groups, no statistically significant differences occurred in the median time to alleviation of all symptoms, nasopharyngeal viral load, or duration of viral shedding. CONCLUSIONS: While MHAA4549A was safe and well tolerated with confirmed exposure, the antibody did not improve clinical outcomes in patients with acute uncomplicated influenza A infection.

The lymphoid cell surface receptor NTB-A: a novel monoclonal antibody target for leukaemia and lymphoma therapeutics.

NTB-A is a CD2-related cell surface protein expressed primarily on lymphoid cells including B-lymphocytes from chronic lymphocytic leukaemia (CLL) and lymphoma patients. We have generated a series of monoclonal antibodies (mAbs) against NTB-A and assessed their therapeutic potential for CLL. Selective mAbs to NTB-A were further tested in functional complement-dependent cytotoxicity (CDC) and antibody-dependent cellular cytotoxicty (ADCC) assays in cell lines and B lymphocytes freshly isolated from CLL patients. While lower levels of NTB-A were detected in T and natural killer (NK) cells, CDC activity was demonstrated primarily in B cells isolated from CLL patients and B lymphoma cell lines. Knockdown of NTB-A by small interfering RNA in target cells results in lower cytotoxicity, demonstrating the specificity of the mAbs. Furthermore, anti NTB-A mAbs demonstrated anti-tumour activity against CA46 human lymphoma xenografts in nude mice and against systemically disseminated Raji human lymphoma cells in severe combined immunodeficient mice. Taken together, these results demonstrate NTB-A as a potential new target for immunotherapy of leukaemia and lymphomas.