Cochlear pharmacokinetics - Micro-computed tomography and learning-prediction modeling for transport parameter determination

The Potential of a CT-Based Machine Learning Radiomics Analysis to Differentiate Brucella and Pyogenic Spondylitis

Background

Pyogenic spondylitis (PS) and Brucella spondylitis (BS) are common spinal infections with similar manifestations, making their differentiation challenging. This study aimed to explore the potential of CT-based radiomics features combined with machine learning algorithms to differentiate PS from BS.

Methods

This retrospective study involved the collection of clinical and radiological information from 138 patients diagnosed with either PS or BS in our hospital between January 2017 and December 2022, based on histopathology examination and/or germ isolations. The region of interest (ROI) was defined by two radiologists using a 3D Slicer open-source platform, utilizing blind analysis of sagittal CT images against histopathological examination results. PyRadiomics, a Python package, was utilized to extract ROI features. Several methods were performed to reduce the dimensionality of the extracted features. Machine learning algorithms were trained and evaluated using techniques like the area under the receiver operating characteristic curve (AUC; confusion matrix-related metrics, calibration plot, and decision curve analysis to assess their ability to differentiate PS from BS. Additionally, permutation feature importance (PFI; local interpretable model-agnostic explanations (LIME; and Shapley additive explanation (SHAP) techniques were utilized to gain insights into the interpretabilities of the models that are otherwise considered opaque black-boxes.

Results

A total of 15 radiomics features were screened during the analysis. The AUC value and Brier score of best the model were 0.88 and 0.13, respectively. The calibration plot and decision curve analysis displayed higher clinical efficiency in the differential diagnosis. According to the interpretation results, the most impactful features on the model output were wavelet LHL small dependence low gray-level emphasis (GLDN).

Conclusion

The CT-based radiomics models that we developed have proven to be useful in reliably differentiating between PS and BS at an early stage and can provide a reliable explanation for the classification results.

Microtechnologies for inner ear drug delivery

PURPOSE OF REVIEW

Treatment of auditory dysfunction is dependent on inner ear drug delivery, with microtechnologies playing an increasingly important role in cochlear access and pharmacokinetic profile control. This review examines recent developments in the field for clinical and animal research environments.

RECENT FINDINGS

Micropump technologies are being developed for dynamic control of flow rates with refillable reservoirs enabling timed delivery of multiple agents for protection or regeneration therapies. These micropumps can be combined with cochlear implants with integral catheters or used independently with cochleostomy or round window membrane (RWM) delivery modalities for therapy development in animal models. Sustained release of steroids with coated cochlear implants remains an active research area with first-time-in-human demonstration of reduced electrode impedances. Advanced coatings containing neurotrophin producing cells have enhanced spiral ganglion neuron survival in animal models, and have proven safe in a human study. Microneedles have emerged for controlled microperforation of the RWM for significant enhancement in permeability, combinable with emerging matrix formulations that optimize biological interaction and drug release kinetics.

SUMMARY

Microsystem technologies are providing enhanced and more controlled access to the inner ear for advanced drug delivery approaches, alone and in conjunction with cochlear implants.

Analysis of Pharmacokinetics in the Cochlea of the Inner Ear

Hearing loss affects >5% of the global population and therefore, has a great social and clinical impact. Sensorineural hearing loss, which can be caused by different factors, such as acoustic trauma, aging, and administration of certain classes of drugs, stems primarily from a dysfunction of the cochlea in the inner ear. Few therapeutic strategies against sensorineural hearing loss are available. To develop effective treatments for this disease, it is crucial to precisely determine the behavior of ototoxic and therapeutic agents in the microenvironment of the cochlea in live animals. Since the 1980s, a number of studies have addressed this issue by different methodologies. However, there is much less information on pharmacokinetics in the cochlea than that in other organs; the delay in ontological pharmacology is likely due to technical difficulties with accessing the cochlea, a tiny organ that is encased with a bony wall and has a fine and complicated internal structure. In this review, we not only summarize the observations and insights obtained in classic and recent studies on pharmacokinetics in the cochlea but also describe relevant analytical techniques, with their strengths, limitations, and prospects.

Inner ear drug delivery through a cochlear implant: Pharmacokinetics in a Macaque experimental model

OBJECTIVES

The method of drug delivery directly into the cochlea with an implantable pump connected to a CI electrode array ensures long-term delivery and effective dose control, and also provides the possibility to use different drugs. The objective is to develop a model of inner ear pharmacokinetics of an implanted cochlea, with the delivery of FITC-Dextran, in the non-human primate model.

DESIGN

A preclinical cochlear electrode array (CI Electrode Array HL14DD, manufactured by Cochlear Ltd.) attached to an implantable peristaltic pump filled with FITC-Dextran was implanted unilaterally in a total of 15 Macaca fascicularis (Mf). Three groups were created (5 Mf in each group), according to three different drug delivery times: 2 hours, 24 hours and 7 days. Perilymph (10 samples, 1μL each) was sampled from the apex of the cochlea and measured immediately after extraction with a spectrofluorometer. After scarifying the specimens, x-Rays and histological analysis were performed.

RESULTS

Surgery, sampling and histological analysis were performed successfully in all specimens. FITC-Dextran quantification showed different patterns, depending on the delivery group. In the 2 hours injection experiment, an increase in FITC-Dextran concentrations over the sample collection time was seen, reaching maximum concentration peaks (420-964µM) between samples 5 and 7, decreasing in successive samples, without returning to baseline. The 24-hours and 7-days injection experiments showed even behaviour throughout the 10 samples obtained, reaching a plateau with mean concentrations ranging from 2144 to 2564 µM and from 1409 to 2502µM, respectively. Statistically significant differences between the 2 hours and 24 hours groups (p = 0.001) and between the 2 hours and 7 days groups (p = 0.037) were observed, while between the 24 hours and 7 days groups no statistical differences were found.

CONCLUSIONS

This experimental study shows that a model of drug delivery and pharmacokinetics using an active pump connected to an electrode array is feasible in Mf. An infusion time ranging from 2 to 24 hours is required to reach a maximum concentration peak at the apex. It establishes then an even concentration profile from base to apex that is maintained throughout the infusion time in Mf. Flow mechanisms during injection and during sampling that may explain such findings may involve cochlear aqueduct flow as well as the possible existence of substance exchange from scala tympani to extracellular spaces, such as the modiolar space or the endolymphatic sinus, acting as a substance reservoir to maintain a relatively flat concentration profile from base to apex during sampling. Leveraging the learnings achieved by experimentation in rodent models, we can move to experiment in non-human primate with the aim of achieving a useful model that provides transferrable data to human pharmacokinetics. Thus, it may broaden clinical and therapeutic approaches to inner ear diseases.

Deep Volumetric Segmentation of Murine Cochlear Compartments from Micro-Computed Tomography Images

Local drug delivery to the inner ear via micropump implants has the potential to be much more effective than oral drug delivery for treating patients with sensorineural hearing loss and to protect hearing from ototoxic insult due to noise exposure or cancer treatments. Designing micropumps to deliver appropriate concentrations of drugs to the necessary cochlear compartments is of paramount importance; however, directly measuring local drug concentrations over time throughout the cochlea is not possible. Recent approaches for indirectly quantifying local drug concentrations in animal models capture a series of magnetic resonance (MR) or micro computed tomography (µCT) images before and after infusion of a contrast agent into the cochlea. These approaches require accurately segmenting important cochlear components (scala tympani (ST), scala media (SM) and scala vestibuli (SV)) in each scan and ensuring that they are registered longitudinally across scans. In this paper, we focus on segmenting cochlear compartments from µCT volumes using V-Net, a convolutional neural network (CNN) architecture for 3-D segmentation. We show that by modifying the V-Net architecture to decrease the numbers of encoder and decoder blocks and to use dilated convolutions enables extracting local estimates of drug concentration that are comparable to those extracted using atlas-based segmentation (3.37%, 4.81%, and 19.65% average relative error in ST, SM, and SV), but in a fraction of the time. We also test the feasibility of training our network on a larger MRI dataset, and then using transfer learning to perform segmentation on a smaller number of µCT volumes, which would enable this technique to be used in the future to characterize drug delivery in the cochlea of larger mammals.

Dose-Dependent Transient Decrease of Impedances by Deep Intracochlear Injection of Triamcinolone With a Cochlear Catheter Prior to Cochlear Implantation-1 Year Data

Administration of low-dose steroids via a catheter inserted into the cochlea to apply pharmaceuticals to more apical regions was previously shown not to be sufficient for long-term reduction of electrode impedances. The aim of the present study was to investigate the effect of intra-cochlear high-dose triamcinolone application on impedances in cochlear implant recipients. Patients received low-dose (4 mg/ml; n = 5) or high-dose (20 mg/ml; n = 5) triamcinolone via a cochlear catheter just prior to the insertion of a Med-El Flex28 electrode. Impedances were measured at defined time points from intra-operatively up to 12 months after first fitting and retrospectively compared with a control group (no steroid application). Patients who received a high-dose application of crystalloid triamcinolone showed significantly reduced impedances in the first fitting measurements compared to the control group. This effect was no longer detectable in patients of the low-dose group at that time. Looking at the different regions of the electrode, the impedance values were lowered significantly only at the basal and medial contacts. At later time points, there were no significant differences between any of the groups. This is the first study to demonstrate a dose-dependent reduction of impedances by deep intra-cochlear injection of triamcinolone in cochlear implant patients. With a high-dose, single application of triamcinolone using a cochlear catheter prior to insertion of a Flex28 electrode, the impedances can be significantly reduced up to and including the first fitting. Although the effect was longer lasting than when compared to low-dose triamcinolone, it was also not permanent.