Intravitreal melphalan hydrochloride vs propylene glycol-free melphalan for retinoblastoma vitreous seeds: Efficacy, toxicity and stability in rabbits models and patients.

The use of intravitreal chemotherapy has revolutionized the treatment of advanced intraocular retinoblastoma, as intravitreal melphalan has enabled difficult-to-treat vitreous tumor seeds to be controlled, leading to many more eyes being saved. However, melphalan hydrochloride (MH) degrades rapidly in solution, increasing logistical complexity with respect to time between medication preparation and administration for intravitreal administration under anesthesia for retinoblastoma. A new propylene glycol-free melphalan (PGFM) formulation has greater stability and could therefore improve access and adoption of intravitreal chemotherapy, allowing more children to retain their eye(s). We compared the efficacy and toxicity of both formulations, using our rabbit xenograft model and clinical patient experience. Three weekly 12.5 µg intravitreal injections of MH or PGFM (right eye), and saline (left eye), were administered to immunosuppressed rabbits harboring human WERI-Rb1 vitreous seed xenografts. Residual live cells were quantified directly, and viability determined by TUNEL staining. Vitreous seeds were reduced 91% by PGFM (p = 0.009), and 88% by MH (p = 0.004; PGFM vs. MH: p = 0.68). All residual cells were TUNEL-positive (non-viable). In separate experiments to assess toxicity, three weekly 12.5 µg injections of MH, PGFM, or saline were administered to non-tumor-bearing rabbits. Serial electroretinography, optical coherence tomography (OCT) and OCT-angiography were performed. PGFM and MH both caused equivalent reductions in electroretinography amplitudes, and loss of retinal microvasculature on OCT-angiography. The pattern of retinal degeneration observed on histopathology suggested that segmental retinal toxicity associated with all melphalan formulations was due to a vitreous concentration gradient-effect. Efficacy and toxicity were assessed for PGFM given immediately (within 1 h of reconstitution) vs. 4 h after reconstitution. Immediate- and delayed-administration of PGFM showed equivalent efficacy and toxicity. In addition, we evaluated efficacy and toxicity in patients (205 eyes) with retinoblastoma vitreous seeds, who were treated with a total of 833 intravitreal injections of either MH or PGFM as standard of care. Of these, we analyzed 118 MH and 131 PGFM monotherapy injections in whom serial ERG measurements were available to model retinal toxicity. Both MH and PGFM caused reductions in electroretinography amplitudes, but with no statistical difference between formulations. Comparing those patient eyes treated exclusively with PGFM versus those treated exclusively with MH, efficacy for tumor control and globe salvage was equivalent (PGFM vs. MH: 96.2% vs. 93.8%, p = 0.56), but PGFM-treated eyes received fewer injections than MH-treated eyes (average 3.2 ± 1.9 vs. 6.4 ± 2.1 injections, p < 0.0001). Taken together, these rabbit experiments and our clinical experience in retinoblastoma patients demonstrate that MH and PGFM have equivalent efficacy and toxicity. PGFM was more stable, with no decreased efficacy or increased toxicity even 4 h after reconstitution. We therefore now use PGFM over traditional MH for our patients for intravitreal treatment of retinoblastoma.

Resident mesenchymal vascular progenitors modulate adaptive angiogenesis and pulmonary remodeling via regulation of canonical Wnt signaling.

Adaptive angiogenesis is necessary for tissue repair, however, it may also be associated with the exacerbation of injury and development of chronic disease. In these studies, we demonstrate that lung mesenchymal vascular progenitor cells (MVPC) modulate adaptive angiogenesis via lineage trace, depletion of MVPC, and modulation of ß-catenin expression. Single cell sequencing confirmed MVPC as multipotential vascular progenitors, thus, genetic depletion resulted in alveolar simplification with reduced adaptive angiogenesis. Following vascular endothelial injury, Wnt activation in MVPC was sufficient to elicit an emphysema-like phenotype characterized by increased MLI, fibrosis, and MVPC driven adaptive angiogenesis. Lastly, activation of Wnt/ß-catenin signaling skewed the profile of human and murine MVPC toward an adaptive phenotype. These data suggest that lung MVPC drive angiogenesis in response to injury and regulate the microvascular niche as well as subsequent distal lung tissue architecture via Wnt signaling.

Photothermal optical coherence tomography of indocyanine green in ex vivo eyes.

Indocyanine green (ICG) is routinely used during surgery to stain the inner limiting membrane (ILM) and provide contrast on white light surgical microscopy. While translation of optical coherence tomography (OCT) for intraoperative imaging during ophthalmic surgery has enhanced visualization, the ILM remains difficult to distinguish from underlying retinal structures and ICG does not provide additional OCT contrast. We present photothermal OCT (PT-OCT) for high-specificity detection of ICG on retinal OCT images. We demonstrate our technique by performing an ILM peel in ex vivo eyes using low ICG concentrations and laser powers. These results establish the feasibility of PT-OCT for intraoperative guidance during retinal surgery.

Assessment of breast pathologies using nonlinear microscopy.

Rapid intraoperative assessment of breast excision specimens is clinically important because up to 40% of patients undergoing breast-conserving cancer surgery require reexcision for positive or close margins. We demonstrate nonlinear microscopy (NLM) for the assessment of benign and malignant breast pathologies in fresh surgical specimens. A total of 179 specimens from 50 patients was imaged with NLM using rapid extrinsic nuclear staining with acridine orange and intrinsic second harmonic contrast generation from collagen. Imaging was performed on fresh, intact specimens without the need for fixation, embedding, and sectioning required for conventional histopathology. A visualization method to aid pathological interpretation is presented that maps NLM contrast from two-photon fluorescence and second harmonic signals to features closely resembling histopathology using hematoxylin and eosin staining. Mosaicking is used to overcome trade-offs between resolution and field of view, enabling imaging of subcellular features over square-centimeter specimens. After NLM examination, specimens were processed for standard paraffin-embedded histology using a protocol that coregistered histological sections to NLM images for paired assessment. Blinded NLM reading by three pathologists achieved 95.4% sensitivity and 93.3% specificity, compared with paraffin-embedded histology, for identifying invasive cancer and ductal carcinoma in situ versus benign breast tissue. Interobserver agreement was κ = 0.88 for NLM and κ = 0.89 for histology. These results show that NLM achieves high diagnostic accuracy, can be rapidly performed on unfixed specimens, and is a promising method for intraoperative margin assessment.

The value of intraoperative optical coherence tomography imaging in vitreoretinal surgery.

PURPOSE OF REVIEW: To evaluate the role of intraoperative optical coherence tomography (iOCT) in vitreoretinal surgery, assess the current state-of-the art, and to examine possible future directions in the field. RECENT FINDINGS: Numerous vitreoretinal surgical conditions and procedures have been described utilizing iOCT. These conditions include macular holes, epiretinal membranes, retinal detachments, and retinopathy of prematurity. Significant alterations appear to occur during surgical manipulations in many of these conditions that can be identified with iOCT. The most common current systems used are portable OCT probes that are either mounted to a microscope or used in a handheld fashion. Prototypes are also being utilized that are integrated into the microscope to allow for true 'real-time' imaging of instrument-tissue interactions. Current generation surgical instrument materials (e.g., metal) limit optimal visualization with integrated OCT systems because of shadowing and light scattering properties. SUMMARY: The role of iOCT in vitreoretinal surgery continues to be defined by active research and enhancements to integrative technologies. Further research is needed to better define the specific applications of iOCT that impact patient outcomes and surgical decision-making. Future advancements in integrative systems, OCT-friendly instrumentation, and software algorithms will further expand the horizon of iOCT in the vitreoretinal surgical theater. As OCT transformed the clinical management of the vitreoretinal conditions, iOCT has the potential to be a paradigm-shifting technology in the operating room.

Swept source optical coherence microscopy using a 1310 nm VCSEL light source.

We demonstrate high speed, swept source optical coherence microscopy (OCM) using a MEMS tunable vertical cavity surface-emitting laser (VCSEL) light source. The light source had a sweep rate of 280 kHz, providing a bidirectional axial scan rate of 560 kHz. The sweep bandwidth was 117 nm centered at 1310 nm, corresponding to an axial resolution of 13.1 µm in air, corresponding to 8.1 µm (9.6 µm spectrally shaped) in tissue. Dispersion mismatch from different objectives was compensated numerically, enabling magnification and field of view to be easily changed. OCM images were acquired with transverse resolutions between 0.86 µm - 3.42 µm using interchangeable 40X, 20X and 10X objectives with ~600 µm x 600 µm, ~1 mm x 1 mm and ~2 mm x 2 mm field-of-view (FOV), respectively. Parasitic variations in path length with beam scanning were corrected numerically. These features enable swept source OCM to be integrated with a wide range of existing scanning microscopes. Large FOV mosaics were generated by serially acquiring adjacent overlapping microscopic fields and combining them in post-processing. Fresh human colon, thyroid and kidney specimens were imaged ex vivo and compared to matching histology sections, demonstrating the ability of OCM to image tissue specimens.

MURIN: Multimodal Retinal Imaging and Navigated-laser-delivery for dynamic and longitudinal tracking of photodamage in murine models.

Laser-induced photodamage is a robust method for investigating retinal pathologies in small animals. However, aiming of the photocoagulation laser is often limited by manual alignment and lacks real-time feedback on lesion location and severity. Here, we demonstrate a multimodality OCT and SLO ophthalmic imaging system with an image-guided scanning laser lesioning module optimized for the murine retina. The proposed system enables targeting of focal and extended area lesions under OCT guidance to benefit visualization of photodamage response and the precision and repeatability of laser lesion models of retinal injury.

Hybrid spiral scanning in a double-clad fiber-based handheld confocal scanning light ophthalmoscope.

High-speed, accessible, and robust in vivo imaging of the human retina is critical for screening of retinal pathologies, such as diabetic retinopathy, age-related macular degeneration, and others. Scanning light ophthalmoscopy (SLO) is a retinal imaging modality that produces digital, en face images of the human retina with superior image gradability rates when compared to the current standard of care in screening for these diseases, namely the flood-illumination handheld fundus camera (HFC). However, current-generation commercial SLO systems are mostly tabletop devices, limiting their accessibility and utility in screening applications. Moreover, most existing SLO systems use raster scan patterns, which are both inefficient and lead to undesired subject gaze drift when used with visible or pseudo-visible illumination. Non-raster scan patterns, especially spiral scanning as described herein, promise advantages in both scan efficiency and reduced subject eye motion. In this work, we introduce a novel "hybrid spiral" scan pattern and the associated hardware design and real-time image reconstruction techniques necessary for its implementation in an SLO system. Building upon this core hybrid spiral scanning SLO (HSS-SLO) technology, we go on to present a complete handheld HSS-SLO system, featuring a fiber-coupled portable patient interface which leverages a dual-clad fiber (DCF) to form a single-path optical topology, thus ensuring mechanically robust co-alignment of illumination and collection apertures, a necessity for a handheld system. The feasibility of HSS-SLO for handheld, in vivo imaging is demonstrated by imaging eight human volunteers.

Structural markers observed with endoscopic 3-dimensional optical coherence tomography correlating with Barrett's esophagus radiofrequency ablation treatment response (with videos).

BACKGROUND: Radiofrequency ablation (RFA) is effective for treating Barrett's esophagus (BE) but often involves multiple endoscopy sessions over several months to achieve complete response. OBJECTIVE: Identify structural markers that correlate with treatment response by using 3-dimensional (3-D) optical coherence tomography (OCT; 3-D OCT). DESIGN: Cross-sectional. SETTING: Single teaching hospital. PATIENTS: Thirty-three patients, 32 male and 1 female, with short-segment (<3 cm) BE undergoing RFA treatment. INTERVENTION: Patients were treated with focal RFA, and 3-D OCT was performed at the gastroesophageal junction before and immediately after the RFA treatment. Patients were re-examined with standard endoscopy 6 to 8 weeks later and had biopsies to rule out BE if not visibly evident. MAIN OUTCOME MEASUREMENTS: The thickness of BE epithelium before RFA and the presence of residual gland-like structures immediately after RFA were determined by using 3-D OCT. The presence of BE at follow-up was assessed endoscopically. RESULTS: BE mucosa was significantly thinner in patients who achieved complete eradication of intestinal metaplasia than in patients who did not achieve complete eradication of intestinal metaplasia at follow-up (257 ± 60 µm vs 403 ± 86 µm; P < .0001). A threshold thickness of 333 µm derived from receiver operating characteristic curves corresponded to a 92.3% sensitivity, 85% specificity, and 87.9% accuracy in predicting the presence of BE at follow-up. The presence of OCT-visible glands immediately after RFA also correlated with the presence of residual BE at follow-up (83.3% sensitivity, 95% specificity, 90.6% accuracy). LIMITATIONS: Single center, cross-sectional study in which only patients with short-segment BE were examined. CONCLUSION: Three-dimensional OCT assessment of BE thickness and residual glands during RFA sessions correlated with treatment response. Three-dimensional OCT may predict responses to RFA or aid in making real-time RFA retreatment decisions in the future.

Characterization of buried glands before and after radiofrequency ablation by using 3-dimensional optical coherence tomography (with videos).

BACKGROUND: Radiofrequency ablation (RFA) is an endoscopic technique used to eradicate Barrett's esophagus (BE). However, such ablation can commonly lead to neosquamous epithelium overlying residual BE glands not visible by conventional endoscopy and may evade detection on random biopsy samples. OBJECTIVE: To demonstrate the capability of endoscopic 3-dimensional optical coherence tomography (3D-OCT) for the identification and characterization of buried glands before and after RFA therapy. DESIGN: Cross-sectional study. SETTING: Single teaching hospital. PATIENTS: Twenty-six male and 1 female white patients with BE undergoing RFA treatment. INTERVENTIONS: 3D-OCT was performed at the gastroesophageal junction in 18 patients before attaining complete eradication of intestinal metaplasia (pre-CE-IM group) and in 16 patients after CE-IM (post-CE-IM group). MAIN OUTCOME MEASUREMENTS: Prevalence, size, and location of buried glands relative to the squamocolumnar junction. RESULTS: 3D-OCT provided an approximately 30 to 60 times larger field of view compared with jumbo and standard biopsy and sufficient imaging depth for detecting buried glands. Based on 3D-OCT results, buried glands were found in 72% of patients (13/18) in the pre-CE-IM group and 63% of patients (10/16) in the post-CE-IM group. The number (mean [standard deviation]) of buried glands per patient in the post-CE-IM group (7.1 [9.3]) was significantly lower compared with the pre-CE-IM group (34.4 [44.6]; P = .02). The buried gland size (P = .69) and distribution (P = .54) were not significantly different before and after CE-IM. LIMITATIONS: A single-center, cross-sectional study comparing patients at different time points in treatment. Lack of 1-to-1 coregistered histology for all OCT data sets obtained in vivo. CONCLUSION: Buried glands were frequently detected with 3D-OCT near the gastroesophageal junction before and after radiofrequency ablation.

Automated instrument-tracking for 4D video-rate imaging of ophthalmic surgical maneuvers.

Intraoperative image-guidance provides enhanced feedback that facilitates surgical decision-making in a wide variety of medical fields and is especially useful when haptic feedback is limited. In these cases, automated instrument-tracking and localization are essential to guide surgical maneuvers and prevent damage to underlying tissue. However, instrument-tracking is challenging and often confounded by variations in the surgical environment, resulting in a trade-off between accuracy and speed. Ophthalmic microsurgery presents additional challenges due to the nonrigid relationship between instrument motion and instrument deformation inside the eye, image field distortion, image artifacts, and bulk motion due to patient movement and physiological tremor. We present an automated instrument-tracking method by leveraging multimodal imaging and deep-learning to dynamically detect surgical instrument positions and re-center imaging fields for 4D video-rate visualization of ophthalmic surgical maneuvers. We are able to achieve resolution-limited tracking accuracy at varying instrument orientations as well as at extreme instrument speeds and image defocus beyond typical use cases. As proof-of-concept, we perform automated instrument-tracking and 4D imaging of a mock surgical task. Here, we apply our methods for specific applications in ophthalmic microsurgery, but the proposed technologies are broadly applicable for intraoperative image-guidance with high speed and accuracy.

Real-time OCT image denoising using a self-fusion neural network.

Optical coherence tomography (OCT) has become the gold standard for ophthalmic diagnostic imaging. However, clinical OCT image-quality is highly variable and limited visualization can introduce errors in the quantitative analysis of anatomic and pathologic features-of-interest. Frame-averaging is a standard method for improving image-quality, however, frame-averaging in the presence of bulk-motion can degrade lateral resolution and prolongs total acquisition time. We recently introduced a method called self-fusion, which reduces speckle noise and enhances OCT signal-to-noise ratio (SNR) by using similarity between from adjacent frames and is more robust to motion-artifacts than frame-averaging. However, since self-fusion is based on deformable registration, it is computationally expensive. In this study a convolutional neural network was implemented to offset the computational overhead of self-fusion and perform OCT denoising in real-time. The self-fusion network was pretrained to fuse 3 frames to achieve near video-rate frame-rates. Our results showed a clear gain in peak SNR in the self-fused images over both the raw and frame-averaged OCT B-scans. This approach delivers a fast and robust OCT denoising alternative to frame-averaging without the need for repeated image acquisition. Real-time self-fusion image enhancement will enable improved localization of OCT field-of-view relative to features-of-interest and improved sensitivity for anatomic features of disease.

Depth-Encoded Spectral Domain Phase Microscopy for Simultaneous Multi-Site Nanoscale Optical Measurements.

Spectral domain phase microscopy (SDPM) is an extension of spectral domain optical coherence tomography (SDOCT) that exploits the extraordinary phase stability of spectrometer-based systems with common-path geometry to resolve sub-wavelength displacements within a sample volume. This technique has been implemented for high resolution axial displacement and velocity measurements in biological samples, but since axial displacement information is acquired serially along the lateral dimension, it has been unable to measure fast temporal dynamics in extended samples. Depth-Encoded SDPM (DESDPM) uses multiple sample arms with unevenly spaced common path reference reflectors to multiplex independent SDPM signals from separate lateral positions on a sample simultaneously using a single interferometer, thereby reducing the time required to detect unique optical events to the integration period of the detector. Here, we introduce DESDPM and demonstrate the ability to acquire useful phase data concurrently at two laterally separated locations in a phantom sample as well as a biological preparation of spontaneously beating chick cardiomyocytes. DESDPM may be a useful tool for imaging fast cellular phenomena such as nervous conduction velocity or contractile motion.

Modeling and optimization of galvanometric point-scanning temporal dynamics.

Galvanometers are ubiquitous in point-scanning applications in optical imaging, display, ranging, manufacturing, and therapeutic technologies. However, galvanometer performance is constrained by finite response times related to mirror size and material properties. We present a model-driven approach for optimizing galvanometer response characteristics by tuning the parameters of the closed-loop galvanometer controller and demonstrate settling time reduction by over 50%. As an imaging proof-of-concept, we implement scan waveforms that take advantage of the optimized galvanometer frequency response to increase linear field-of-view, signal-to-noise ratio, contrast-to-noise ratio, and speed. The hardware methods presented may be directly implemented on galvanometer controllers without the need for specialized equipment and used in conjunction with customized scan waveforms to further optimize scanning performance.

Advances in multimodal imaging in ophthalmology.

Multimodality ophthalmic imaging systems aim to enhance the contrast, resolution, and functionality of existing technologies to improve disease diagnostics and therapeutic guidance. These systems include advanced acquisition and post-processing methods using optical coherence tomography (OCT), combined scanning laser ophthalmoscopy and OCT systems, adaptive optics, surgical guidance, and photoacoustic technologies. Here, we provide an overview of these ophthalmic imaging systems and their clinical and basic science applications.

LIFE: A Generalizable Autodidactic Pipeline for 3D OCT-A Vessel Segmentation.

Optical coherence tomography (OCT) is a non-invasive imaging technique widely used for ophthalmology. It can be extended to OCT angiography (OCT-A), which reveals the retinal vasculature with improved contrast. Recent deep learning algorithms produced promising vascular segmentation results; however, 3D retinal vessel segmentation remains difficult due to the lack of manually annotated training data. We propose a learning-based method that is only supervised by a self-synthesized modality named local intensity fusion (LIF). LIF is a capillary-enhanced volume computed directly from the input OCT-A. We then construct the local intensity fusion encoder (LIFE) to map a given OCT-A volume and its LIF counterpart to a shared latent space. The latent space of LIFE has the same dimensions as the input data and it contains features common to both modalities. By binarizing this latent space, we obtain a volumetric vessel segmentation. Our method is evaluated in a human fovea OCT-A and three zebrafish OCT-A volumes with manual labels. It yields a Dice score of 0.7736 on human data and 0.8594 ± 0.0275 on zebrafish data, a dramatic improvement over existing unsupervised algorithms.

Controlling Droplet Marangoni Flows to Improve Microscopy-Based TB Diagnosis.

In developing countries, the most common diagnostic method for tuberculosis (TB) is microscopic examination sputum smears. Current assessment requires time-intensive inspection across the microscope slide area, and this contributes to its poor diagnostic sensitivity of ≈50%. Spatially concentrating TB bacteria in a smaller area is one potential approach to improve visual detection and potentially increase sensitivity. We hypothesized that a combination of magnetic concentration and induced droplet Marangoni flow would spatially concentrate Mycobacterium tuberculosis on the slide surface by preferential deposition of beads and TB-bead complexes in the center of an evaporating droplet. To this end, slide substrate and droplet solvent thermal conductivities and solvent surface tension, variables known to impact microfluidic flow patterns in evaporating droplets, were varied to select the most appropriate slide surface coating. Optimization in a model system used goniometry, optical coherence tomography, and microscope images of the final deposition pattern to observe the droplet flows and maximize central deposition of 1 µm fluorescent polystyrene particles and 200 nm nanoparticles (NPs) in 2 µL droplets. Rain-X® polysiloxane glass coating was identified as the best substrate material, with a PBS-Tween droplet solvent. The use of smaller, 200 nm magnetic NPs instead of larger 1 µm beads allowed for bright field imaging of bacteria. Using these optimized components, we compared standard smear methods to the Marangoni-based spatial concentration system, which was paired with magnetic enrichment using iron oxide NPs, isolating M. bovis BCG (BCG) from samples containing 0 and 103 to 106 bacilli/mL. Compared to standard smear preparation, paired analysis demonstrated a combined volumetric and spatial sample enrichment of 100-fold. With further refinement, this magnetic/Marangoni flow concentration approach is expected to improve whole-pathogen microscopy-based diagnosis of TB and other infectious diseases.

Intravitreal HDAC Inhibitor Belinostat Effectively Eradicates Vitreous Seeds Without Retinal Toxicity In Vivo in a Rabbit Retinoblastoma Model.

Purpose: Current melphalan-based regimens for intravitreal chemotherapy for retinoblastoma vitreous seeds are effective but toxic to the retina. Thus, alternative agents are needed. Based on the known biology of histone deacetylases (HDACs) in the retinoblastoma pathway, we systematically studied whether the HDAC inhibitor belinostat is a viable, molecularly targeted alternative agent for intravitreal delivery that might provide comparable efficacy, without toxicity. Methods: In vivo pharmacokinetic experiments in rabbits and in vitro cytotoxicity experiments were performed to determine the 90% inhibitory concentration (IC90). Functional toxicity by electroretinography and structural toxicity by optical coherence tomography (OCT), OCT angiography, and histopathology were evaluated in rabbits following three injections of belinostat 350 µg (2× IC90) or 700 µg (4× IC90), compared with melphalan 12.5 µg (rabbit equivalent of the human dose). The relative efficacy of intravitreal belinostat versus melphalan to treat WERI-Rb1 human cell xenografts in rabbit eyes was directly quantified. RNA sequencing was used to assess belinostat-induced changes in RB cell gene expression. Results: The maximum nontoxic dose of belinostat was 350 µg, which caused no reductions in electroretinography parameters, retinal microvascular loss on OCT angiography, or retinal degeneration. Melphalan caused severe retinal structural and functional toxicity. Belinostat 350 µg (equivalent to 700 µg in the larger human eye) was equally effective at eradicating vitreous seeds in the rabbit xenograft model compared with melphalan (95.5% reduction for belinostat, P < 0.001; 89.4% reduction for melphalan, P < 0.001; belinostat vs. melphalan, P = 0.10). Even 700 µg belinostat (equivalent to 1400 µg in humans) caused only minimal toxicity. Widespread changes in gene expression resulted. Conclusions: Molecularly targeted inhibition of HDACs with intravitreal belinostat was equally effective as standard-of-care melphalan but without retinal toxicity. Belinostat may therefore be an attractive agent to pursue clinically for intravitreal treatment of retinoblastoma.

Efficacy, Toxicity, and Pharmacokinetics of Intra-Arterial Chemotherapy Versus Intravenous Chemotherapy for Retinoblastoma in Animal Models and Patients.

Purpose: Through controlled comparative rabbit experiments and parallel patient studies, our purpose was to understand mechanisms underlying differences in efficacy and toxicity between intra-arterial chemotherapy (IAC) and intravenous chemotherapy (IVC). Methods: In rabbits, ocular tissue drug levels were measured following IAC and IVC. Retinal toxicity was assessed using electroretinography, fluorescein angiography, optical coherence tomography (OCT) and OCT angiography. Efficacy to eradicate retinoblastoma orthotopic xenografts was compared. In IAC and IVC patients, we measured blood carboplatin pharmacokinetics and compared efficacy and toxicity. Results: In rabbits receiving IAC, maximum carboplatin levels were 134 times greater in retina (P = 0.01) and 411 times greater in vitreous (P < 0.001), and total carboplatin (area under the curve) was 123 times greater in retina (P = 0.005) and 131 times greater in vitreous (P = 0.02) compared with IVC. Melphalan levels were 12 times greater (P = 0.003) in retina and 26 times greater in vitreous (P < 0.001) for IAC. Blood levels were not different. IAC melphalan (but not IV melphalan or IV carboplatin, etoposide, and vincristine) caused widespread apoptosis in retinoblastoma xenografts but no functional retinal toxicity or cytopenias. In patients, blood levels following IVC were greater (P < 0.001) but, when adjusted for treatment dose, were not statistically different. Per treatment cycle in patients, IVC caused higher rates of anemia (0.32 ± 0.29 vs. 0.01 ± 0.04; P = 0.0086), thrombocytopenia (0.5 ± 0.42 vs. 0.0 ± 0.0; P = 0.0042), and neutropenia (0.58 ± 0.3 vs. 0.31 ± 0.25; P = 0.032) but lower treatment success rates (P = 0.0017). Conclusions: The greater efficacy and lower systemic toxicity with IAC appear to be attributable to the greater ocular-to-systemic drug concentration ratio compared with IVC. Translational Relevance: Provides an overarching hypothesis for a mechanism of efficacy/toxicity to guide future drug development.