Effectiveness of 225Ac-Labeled Anti-EGFR Radioimmunoconjugate in EGFR-Positive Kirsten Rat Sarcoma Viral Oncogene and BRAF Mutant Colorectal Cancer Models.

Eighty percent of colorectal cancers (CRCs) overexpress epidermal growth factor receptor (EGFR). Kirsten rat sarcoma viral oncogene (KRAS) mutations are present in 40% of CRCs and drive de novo resistance to anti-EGFR drugs. BRAF oncogene is mutated in 7%-10% of CRCs, with even worse prognosis. We have evaluated the effectiveness of [225Ac]Ac-macropa-nimotuzumab in KRAS mutant and in KRAS wild-type and BRAFV600E mutant EGFR-positive CRC cells in vitro and in vivo. Anti-CD20 [225Ac]Ac-macropa-rituximab was developed and used as a nonspecific radioimmunoconjugate. Methods: Anti-EGFR antibody nimotuzumab was radiolabeled with 225Ac via an 18-membered macrocyclic chelator p-SCN-macropa. The immunoconjugate was characterized using flow cytometry, radioligand binding assay, and high-performance liquid chromatography, and internalization was studied using live-cell imaging. In vitro cytotoxicity was evaluated in 2-dimensional monolayer EGFR-positive KRAS mutant DLD-1, SW620, and SNU-C2B; in KRAS wild-type and BRAFV600E mutant HT-29 CRC cell lines; and in 3-dimensional spheroids. Dosimetry was studied in healthy mice. The in vivo efficacy of [225Ac]Ac-macropa-nimotuzumab was evaluated in mice bearing DLD-1, SW620, and HT-29 xenografts after treatment with 3 doses of 13 kBq/dose administered 10 d apart. Results: In all cell lines, in vitro studies showed enhanced cytotoxicity of [225Ac]Ac-macropa-nimotuzumab compared with nimotuzumab and controls. The inhibitory concentration of 50% in the DLD-1 cell line was 1.8 nM for [225Ac]Ac-macropa-nimotuzumab versus 84.1 nM for nimotuzumab. Similarly, the inhibitory concentration of 50% was up to 79-fold lower for [225Ac]Ac-macropa-nimotuzumab than for nimotuzumab in KRAS mutant SNU-C2B and SW620 and in KRAS wild-type and BRAFV600E mutant HT-29 CRC cell lines. A similar trend was observed for 3-dimensional spheroids. Internalization peaked 24-48 h after incubation and depended on EGFR expression. In the [225Ac]Ac-macropa-nimotuzumab group, 3 of 7 mice bearing DLD-1 tumors had complete remission. Median survival was 40 and 34 d for mice treated with phosphate-buffered saline and [225Ac]Ac-macropa-rituximab (control), respectively, whereas it was not reached for the [225Ac]Ac-macropa-nimotuzumab group (>90 d). Similarly, median survival of mice bearing HT-29 xenografts was 16 and 12.5 d for those treated with [225Ac]Ac-macropa-rituximab and phosphate-buffered saline, respectively, and was not reached for those treated with [225Ac]Ac-macropa-nimotuzumab (>90 d). One of 7 mice bearing HT-29 xenografts and treated using [225Ac]Ac-macropa-nimotuzumab had complete remission. Compared with untreated mice, [225Ac]Ac-macropa-nimotuzumab more than doubled (16 vs. 41 d) the median survival of mice bearing SW620 xenografts. Conclusion: [225Ac]Ac-macropa-nimotuzumab is effective against KRAS mutant and BRAFV600E mutant CRC models.

Hybrid Topological Data Analysis and Deep Learning for Basal Cell Carcinoma Diagnosis.

A critical clinical indicator for basal cell carcinoma (BCC) is the presence of telangiectasia (narrow, arborizing blood vessels) within the skin lesions. Many skin cancer imaging processes today exploit deep learning (DL) models for diagnosis, segmentation of features, and feature analysis. To extend automated diagnosis, recent computational intelligence research has also explored the field of Topological Data Analysis (TDA), a branch of mathematics that uses topology to extract meaningful information from highly complex data. This study combines TDA and DL with ensemble learning to create a hybrid TDA-DL BCC diagnostic model. Persistence homology (a TDA technique) is implemented to extract topological features from automatically segmented telangiectasia as well as skin lesions, and DL features are generated by fine-tuning a pre-trained EfficientNet-B5 model. The final hybrid TDA-DL model achieves state-of-the-art accuracy of 97.4% and an AUC of 0.995 on a holdout test of 395 skin lesions for BCC diagnosis. This study demonstrates that telangiectasia features improve BCC diagnosis, and TDA techniques hold the potential to improve DL performance.

Basal Cell Carcinoma Diagnosis with Fusion of Deep Learning and Telangiectasia Features.

In recent years, deep learning (DL) has been used extensively and successfully to diagnose different cancers in dermoscopic images. However, most approaches lack clinical inputs supported by dermatologists that could aid in higher accuracy and explainability. To dermatologists, the presence of telangiectasia, or narrow blood vessels that typically appear serpiginous or arborizing, is a critical indicator of basal cell carcinoma (BCC). Exploiting the feature information present in telangiectasia through a combination of DL-based techniques could create a pathway for both, improving DL results as well as aiding dermatologists in BCC diagnosis. This study demonstrates a novel "fusion" technique for BCC vs non-BCC classification using ensemble learning on a combination of (a) handcrafted features from semantically segmented telangiectasia (U-Net-based) and (b) deep learning features generated from whole lesion images (EfficientNet-B5-based). This fusion method achieves a binary classification accuracy of 97.2%, with a 1.3% improvement over the corresponding DL-only model, on a holdout test set of 395 images. An increase of 3.7% in sensitivity, 1.5% in specificity, and 1.5% in precision along with an AUC of 0.99 was also achieved. Metric improvements were demonstrated in three stages: (1) the addition of handcrafted telangiectasia features to deep learning features, (2) including areas near telangiectasia (surround areas), (3) discarding the noisy lower-importance features through feature importance. Another novel approach to feature finding with weak annotations through the examination of the surrounding areas of telangiectasia is offered in this study. The experimental results show state-of-the-art accuracy and precision in the diagnosis of BCC, compared to three benchmark techniques. Further exploration of deep learning techniques for individual dermoscopy feature detection is warranted.

LAMA: Lesion-Aware Mixup Augmentation for Skin Lesion Segmentation.

Deep learning can exceed dermatologists' diagnostic accuracy in experimental image environments. However, inaccurate segmentation of images with multiple skin lesions can be seen with current methods. Thus, information present in multiple-lesion images, available to specialists, is not retrievable by machine learning. While skin lesion images generally capture a single lesion, there may be cases in which a patient's skin variation may be identified as skin lesions, leading to multiple false positive segmentations in a single image. Conversely, image segmentation methods may find only one region and may not capture multiple lesions in an image. To remedy these problems, we propose a novel and effective data augmentation technique for skin lesion segmentation in dermoscopic images with multiple lesions. The lesion-aware mixup augmentation (LAMA) method generates a synthetic multi-lesion image by mixing two or more lesion images from the training set. We used the publicly available International Skin Imaging Collaboration (ISIC) 2017 Challenge skin lesion segmentation dataset to train the deep neural network with the proposed LAMA method. As none of the previous skin lesion datasets (including ISIC 2017) has considered multiple lesions per image, we created a new multi-lesion (MuLe) segmentation dataset utilizing publicly available ISIC 2020 skin lesion images with multiple lesions per image. MuLe was used as a test set to evaluate the effectiveness of the proposed method. Our test results show that the proposed method improved the Jaccard score 8.3% from 0.687 to 0.744 and the Dice score 5% from 0.7923 to 0.8321 over a baseline model on MuLe test images. On the single-lesion ISIC 2017 test images, LAMA improved the baseline model's segmentation performance by 0.08%, raising the Jaccard score from 0.7947 to 0.8013 and the Dice score 0.6% from 0.8714 to 0.8766. The experimental results showed that LAMA improved the segmentation accuracy on both single-lesion and multi-lesion dermoscopic images. The proposed LAMA technique warrants further study.

Apoptotic Factors and Mitochondrial Complexes Assist Determination of Strain-Specific Susceptibility of Mice to Parkinsonian Neurotoxin MPTP.

Identification of genetic mutations in Parkinson's disease (PD) promulgates the genetic nature of disease susceptibility. Resilience-associated genes being unknown till date, the normal genetic makeup of an individual may be determinative too. Our earlier studies comparing the substantia nigra (SN) and striatum of C57BL/6J, CD-1 mice, and their F1-crossbreds demonstrated the neuroprotective role of admixing against the neurotoxin MPTP. Furthermore, the differences in levels of mitochondrial fission/fusion proteins in the SN of parent strains imply effects on mitochondrial biogenesis. Our present investigations suggest that the baseline levels of apoptotic factors Bcl-2, Bax, and AIF differ across the three strains and are differentially altered in SN following MPTP administration. The reduction in complex-I levels exclusively in MPTP-injected C57BL/6J reiterates mitochondrial involvement in PD pathogenesis. The MPTP-induced increase in complex-IV, in the nigra of both parent strains, may be compensatory in nature. The ultrastructural evaluation showed fairly preserved mitochondria in the dopaminergic neurons of CD-1 and F1-crossbreds. However, in CD-1, the endoplasmic reticulum demonstrated distinct luminal enlargement, bordering onto ballooning, suggesting proteinopathy as a possible initial trigger.The increase in α-synuclein in the pars reticulata of crossbreds suggests a supportive role for this output nucleus in compensating for the lost function of pars compacta. Alternatively, since α-synuclein over-expression occurs in different brain regions in PD, the α-synuclein increase here may suggest a similar pathogenic outcome. Further understanding is required to resolve this biological contraption. Nevertheless, admixing reduces the risk to MPTP by favoring anti-apoptotic consequences. Similar neuroprotection may be envisaged in the admixed populace of Anglo-Indians.

Skin Lesion Segmentation in Dermoscopic Images with Noisy Data.

We propose a deep learning approach to segment the skin lesion in dermoscopic images. The proposed network architecture uses a pretrained EfficientNet model in the encoder and squeeze-and-excitation residual structures in the decoder. We applied this approach on the publicly available International Skin Imaging Collaboration (ISIC) 2017 Challenge skin lesion segmentation dataset. This benchmark dataset has been widely used in previous studies. We observed many inaccurate or noisy ground truth labels. To reduce noisy data, we manually sorted all ground truth labels into three categories - good, mildly noisy, and noisy labels. Furthermore, we investigated the effect of such noisy labels in training and test sets. Our test results show that the proposed method achieved Jaccard scores of 0.807 on the official ISIC 2017 test set and 0.832 on the curated ISIC 2017 test set, exhibiting better performance than previously reported methods. Furthermore, the experimental results showed that the noisy labels in the training set did not lower the segmentation performance. However, the noisy labels in the test set adversely affected the evaluation scores. We recommend that the noisy labels should be avoided in the test set in future studies for accurate evaluation of the segmentation algorithms.

Biparatopic anti-HER2 drug radioconjugates as breast cancer theranostics.

BACKGROUND: HER2 is overexpressed in 25-30% of breast cancer. Multiple domains targeting of a receptor can have synergistic/additive therapeutic effects. METHODS: Two domain-specific ADCs trastuzumab-PEG6-DM1 (domain IV) and pertuzumab-PEG6-DM1 (domain II) were developed, characterised and radiolabeled to obtain [89Zr]Zr-trastuzumab-PEG6-DM1 and [67Cu]Cu-pertuzumab-PEG6-DM1 to study their in vitro (binding assay, internalisation and cytotoxicity) and in vivo (pharmacokinetics, biodistribution and immunoPET/SPECT imaging) characteristics. RESULTS: The ADCs had an average drug-to-antibody ratio of 3. Trastuzumab did not compete with [67Cu]Cu-pertuzumab-PEG6-DM1 for binding to HER2. The highest antibody internalisation was observed with the combination of ADCs in BT-474 cells compared with single antibodies or ADCs. The combination of the two ADCs had the lowest IC50 compared with treatment using the single ADCs or controls. Pharmacokinetics showed biphasic half-lives with fast distribution and slow elimination, and an AUC that was five-fold higher for [89Zr]Zr-trastuzumab-PEG6-DM1 compared with [67Cu]Cu-pertuzumab-PEG6-DM1. Tumour uptake of [89Zr]Zr-trastuzumab-PEG6-DM1 was 51.3 ± 17.3% IA/g (BT-474), and 12.9 ± 2.1% IA/g (JIMT-1) which was similarly to [67Cu]Cu-pertuzumab-PEG6-DM1. Mice pre-blocked with pertuzumab had [89Zr]Zr-trastuzumab-PEG6-DM1 tumour uptakes of 66.3 ± 33.9% IA/g (BT-474) and 25.3 ± 4.9% IA/g (JIMT-1) at 120 h p.i. CONCLUSION: Using these biologics simultaneously as biparatopic theranostic agents has additive benefits.

Identification of Volatile Organic Liquids by Combining an Array of Fiber-Optic Sensors and Machine Learning.

In this paper, we report an array of fiber-optic sensors based on the Fabry-Perot interference principle and machine learning-based analyses for identifying volatile organic liquids (VOLs). Three optical fiber tip sensors with different surfaces were included in the array of sensors to improve the accuracy for identifying liquids: an intrinsic (unmodified) flat cleaved endface, a hydrophobic-coated endface, and a hydrophilic-coated endface. The time-transient responses of evaporating droplets from the optical fiber tip sensors were monitored and collected following the controlled immersion tests of 11 different organic liquids. A continuous wavelet transform was used to convert the time-transient response signal into images. These images were then utilized to train convolution neural networks for classification (identification of VOLs). We show that diversity in the information collected using the array of three sensors helps machine learning-based methods perform significantly better. We explore different pipelines for combining the information from the array of sensors within a machine learning framework and their effect on the robustness of models. The results showed that the machine learning-based methods achieved high accuracy in their classification of different liquids based on their droplet evaporation time-transient events.

Improving Automatic Melanoma Diagnosis Using Deep Learning-Based Segmentation of Irregular Networks.

Deep learning has achieved significant success in malignant melanoma diagnosis. These diagnostic models are undergoing a transition into clinical use. However, with melanoma diagnostic accuracy in the range of ninety percent, a significant minority of melanomas are missed by deep learning. Many of the melanomas missed have irregular pigment networks visible using dermoscopy. This research presents an annotated irregular network database and develops a classification pipeline that fuses deep learning image-level results with conventional hand-crafted features from irregular pigment networks. We identified and annotated 487 unique dermoscopic melanoma lesions from images in the ISIC 2019 dermoscopic dataset to create a ground-truth irregular pigment network dataset. We trained multiple transfer learned segmentation models to detect irregular networks in this training set. A separate, mutually exclusive subset of the International Skin Imaging Collaboration (ISIC) 2019 dataset with 500 melanomas and 500 benign lesions was used for training and testing deep learning models for the binary classification of melanoma versus benign. The best segmentation model, U-Net++, generated irregular network masks on the 1000-image dataset. Other classical color, texture, and shape features were calculated for the irregular network areas. We achieved an increase in the recall of melanoma versus benign of 11% and in accuracy of 2% over DL-only models using conventional classifiers in a sequential pipeline based on the cascade generalization framework, with the highest increase in recall accompanying the use of the random forest algorithm. The proposed approach facilitates leveraging the strengths of both deep learning and conventional image processing techniques to improve the accuracy of melanoma diagnosis. Further research combining deep learning with conventional image processing on automatically detected dermoscopic features is warranted.

ChimeraNet: U-Net for Hair Detection in Dermoscopic Skin Lesion Images.

Hair and ruler mark structures in dermoscopic images are an obstacle preventing accurate image segmentation and detection of critical network features. Recognition and removal of hairs from images can be challenging, especially for hairs that are thin, overlapping, faded, or of similar color as skin or overlaid on a textured lesion. This paper proposes a novel deep learning (DL) technique to detect hair and ruler marks in skin lesion images. Our proposed ChimeraNet is an encoder-decoder architecture that employs pretrained EfficientNet in the encoder and squeeze-and-excitation residual (SERes) structures in the decoder. We applied this approach at multiple image sizes and evaluated it using the publicly available HAM10000 (ISIC2018 Task 3) skin lesion dataset. Our test results show that the largest image size (448 × 448) gave the highest accuracy of 98.23 and Jaccard index of 0.65 on the HAM10000 (ISIC 2018 Task 3) skin lesion dataset, exhibiting better performance than for two well-known deep learning approaches, U-Net and ResUNet-a. We found the Dice loss function to give the best results for all measures. Further evaluated on 25 additional test images, the technique yields state-of-the-art accuracy compared to 8 previously reported classical techniques. We conclude that the proposed ChimeraNet architecture may enable improved detection of fine image structures. Further application of DL techniques to detect dermoscopy structures is warranted.

Engineering of a Fully Human Anti-MUC-16 Antibody and Evaluation as a PET Imaging Agent.

Antibodies that recognize cancer biomarkers, such as MUC16, can be used as vehicles to deliver contrast agents (imaging) or cytotoxic payloads (therapy) to the site of tumors. MUC16 is overexpressed in 80% of epithelial ovarian cancer (EOC) and 65% of pancreatic ductal adenocarcinomas (PDAC), where effective 'theranostic' probes are much needed. This work aims to develop fully human antibodies against MUC16 and evaluate them as potential immuno-PET imaging probes for detecting ovarian and pancreatic cancers. We developed a fully human monoclonal antibody, M16Ab, against MUC16 using phage display. M16Ab was conjugated with p-SCN-Bn-DFO and radiolabeled with 89Zr. 89Zr-DFO-M16Ab was then evaluated for binding specificity and affinity using flow cytometry. In vivo evaluation of 89Zr-DFO-M16Ab was performed by microPET/CT imaging at different time points at 24−120 h post injection (p.i.) and ex vivo biodistribution studies in mice bearing MUC16-expressing OVCAR3, SKOV3 (ovarian) and SW1990 (pancreatic) xenografts. 89Zr-DFO-M16Ab bound specifically to MUC16-expressing cancer cells with an EC50 of 10nM. 89Zr-DFO-M16Ab was stable in serum and showed specific uptake and retention in tumor xenografts even after 120 h p.i. (microPET/CT) with tumor-to-blood ratios > 43 for the SW1990 xenograft. Specific tumor uptake was observed for SW1990/OVCAR3 xenografts but not in MUC16-negative SKOV3 xenografts. Pharmacokinetic study shows a relatively short distribution (t1/2α) and elimination half-life (t1/2ß) of 4.4 h and 99 h, respectively. In summary, 89Zr-DFO-M16Ab is an effective non-invasive imaging probe for ovarian and pancreatic cancers and shows promise for further development of theranostic radiopharmaceuticals.

Simultaneous Imaging and Therapy Using Epitope-Specific Anti-Epidermal Growth Factor Receptor (EGFR) Antibody Conjugates.

Matuzumab and nimotuzumab are anti-EGFR monoclonal antibodies that bind to different epitopes of domain III of EGFR. We developed 89Zr-matuzumab as a PET probe for diagnosis/monitoring of response to treatment of a noncompeting anti-EGFR nimotuzumab antibody drug conjugate (ADC) using mouse colorectal cancer (CRC) xenografts. We developed 89Zr-matuzumab and performed quality control in EGFR-positive DLD-1 cells. The KD of matuzumab, DFO-matuzumab and 89Zr-matuzumab in DLD-1 cells was 5.9, 6.2 and 3 nM, respectively. A competitive radioligand binding assay showed that 89Zr-matuzumab and nimotuzumab bound to noncompeting epitopes of EGFR. MicroPET/CT imaging and biodistribution of 89Zr-matuzumab in mice bearing EGFR-positive xenografts (HT29, DLD-1 and MDA-MB-231) showed high uptake that was blocked with pre-dosing with matuzumab but not with the noncompeting binder nimotuzumab. We evaluated nimotuzumab-PEG6-DM1 ADC in CRC cells. IC50 of nimotuzumab-PEG6-DM1 in SNU-C2B, DLD-1 and SW620 cells was dependent on EGFR density and was up to five-fold lower than that of naked nimotuzumab. Mice bearing the SNU-C2B xenograft were treated using three 15 mg/kg doses of nimotuzumab-PEG6-DM1, and 89Zr-matuzumab microPET/CT was used to monitor the response to treatment. Treatment resulted in complete remission of the SNU-C2B tumor in 2/3 mice. Matuzumab and nimotuzumab are noncompeting and can be used simultaneously.

Machine learning identifies liquids employing a simple fiber-optic tip sensor.

We proposed an extremely simple fiber-optic tip sensor system to identify liquids by combining their corresponding droplet evaporation events with analyses using machine learning techniques. Pendant liquid droplets were suspended from the cleaved endface of a single-mode fiber during the experiment. The optical fiber-droplet interface and the droplet-air interface served as two partial reflectors of an extrinsic Fabry-Perot interferometer (EFPI) with a liquid droplet cavity. As the liquid pendant droplet evaporated, its length diminished. A light source can be used to observe the effective change in the net reflectivity of the optical fiber sensor system by observing the resulting optical interference phenomenon of the reflected waves. Using a single-wavelength probing light source, the entire evaporation event of the liquid droplet was precisely captured. The measured time transient response from the fiber-optic tip sensor to an evaporation event of a liquid droplet of interest was then transformed into image data using a continuous wavelet transform. The obtained image data was used to fine-tune pre-trained convolution neural networks (CNNs) for the given task. The results demonstrated that machine learning-based classification methods achieved greater than 98% accuracy in classifying different liquids based on their corresponding droplet evaporation processes, measured by the fiber-optic tip sensor.

Identification and Mechanistic Characterization of a Peptide Inhibitor of Glycogen Synthase Kinase (GSK3ß) Derived from the Disrupted in Schizophrenia 1 (DISC1) Protein.

Glycogen synthase kinase 3-beta (GSK3ß) is a critical regulator of several cellular pathways involved in neurodevelopment and neuroplasticity and as such is a potential focus for the discovery of new neurotherapeutics toward the treatment of neuropsychiatric and neurodegenerative diseases. The majority of efforts to develop inhibitors of GSK3ß have been focused on developing small molecule inhibitors that compete with adenosine triphosphate (ATP) through direct interaction with the ATP binding site. This strategy has presented selectivity challenges due to the evolutionary conservation of this domain within the kinome. The disrupted in schizophrenia 1 (DISC1) protein has previously been shown to bind and inhibit GSK3ß activity. Here, we report the characterization of a 44-mer peptide derived from human DISC1 (hDISCtide) that is sufficient to both bind and inhibit GSK3ß in a noncompetitive mode distinct from classical ATP competitive inhibitors. Based on multiple independent biochemical and biophysical assays, we propose that hDISCtide interacts at two distinct regions of GSK3ß: an inhibitory region that partially overlaps with the binding site of FRATide, a well-known GSK3ß binding peptide, and a specific binding region that is unique to hDISCtide. Taken together, our findings present a novel avenue for developing a peptide-based selective inhibitor of GSK3ß.