In vivo measurement of light penetration in mouse and dog heads and human-head simulation for photocatalytic therapy in Alzheimer disease

Technical Summary: In Vivo Measurement of Light Penetration in Mouse and Dog Heads and Human-Head Simulation for Photocatalytic Therapy in Alzheimer Disease

Problem Statement
The authors address a critical translational bottleneck in developing photocatalytic therapy for Alzheimer's disease (AD). While their previous work demonstrated that a photocatalyst binding to amyloid-β (Aβ) fibrils, combined with light irradiation, could clear Aβ in mouse brains through the parietal scalp, it remains unclear if sufficient light energy can penetrate the thicker skulls and tissues of larger animals, such as humans. Existing optical penetration simulations often rely on parameters derived from excised tissues or different wavelength ranges (800–900 nm), which may not accurately reflect the physiological conditions or the specific excitation wavelengths (500–700 nm) required for their small-molecule photocatalyst. Consequently, there is a lack of validated data to determine if transcranial light delivery can achieve the energy thresholds necessary for therapeutic photo-oxygenation in humans.

Methodology
The study employed a multi-stage approach combining in vivo (and post-mortem) measurements with Monte Carlo simulations:

1. Experimental Setup:

   • Light Source: An orange LED system (center wavelength 610 nm, FWHM 86 nm) covering the 500–750 nm range.
   • Measurement Device: A needle-shaped probe with a side detector window connected to a spectrophotometer. Probes of varying diameters (0.8 mm for mice, 4.0 mm for dogs) were used to minimize tissue disruption.
   • Subjects: 8-week-old male C57BL/6J mice (anesthetized) and Beagle dogs (post-mortem specimens obtained within 1 hour of euthanasia).
   • Procedure: In mice, the probe was inserted into the hippocampus region via a craniotomy. In dogs, measurements were taken sequentially for three distinct layers: (1) skin (parietal scalp surface to inner scalp), (2) skull (inner scalp to inner skull), and (3) brain (brain surface to hippocampus vicinity). Spectral transmittance ($T$) was calculated by comparing the intensity at the detector to the baseline intensity directly beneath the light source.

2. Simulation and Parameter Optimization:

   • Model: A three-layer planar scattering medium model (skin, skull, brain) was constructed using Advanced Systems Analysis Program (ASAP) software.
   • Parameter Fitting: Initial optical parameters (scattering coefficient $\mu_s$ and absorption coefficient $\mu_a$) were sourced from literature and software toolkits. These were modified to fit the experimental transmittance data using three approaches:
     • Scaling: Uniform adjustment of coefficients.
     • Mixing: Interpolation between tissue components (e.g., white and gray matter).
     • Addition: Incorporating specific chromophore absorption (e.g., hemoglobin).
   • Validation: The modified parameters were validated against independent mouse data and then applied to a scaled human head model to estimate deep-tissue penetration.

Key Results

• Experimental Transmittance: Spectral transmittance in both species showed rapid changes in the 580–620 nm range, attributed to hemoglobin absorption. In dogs, transmittance to the hippocampus dropped below $10^{-4}$ in most regions, with significant noise in the short-wavelength range due to low signal intensity.
• Optical Parameter Modification:
  • Skin: Initial simulations overestimated transmittance; scaling the scattering coefficient corrected this.
  • Skull: Initial simulations underestimated transmittance; scaling both scattering and absorption coefficients improved the fit.
  • Brain: A combination of mixing gray/white matter parameters and adding hemoglobin absorption at 580 nm was required to reproduce the rapid transmittance drop observed in measurements.
• Human Simulation:
  • Attenuation Profile: The human skin layer (approx. 3x thicker than in dogs) was identified as the primary source of light attenuation, particularly below 580 nm. The skull layer caused relatively modest attenuation compared to the skin.
  • Wavelength Dependence: At 620 nm, approximately $10^{-7}$ of incident light reached the brain surface; at 680 nm, this increased to $10^{-6}$.
  • Depth Penetration: For wavelengths >650 nm, the penetration depth for a 10-fold attenuation was estimated at ~1.5 mm in skin, but ~5.5 mm in the skull and brain layers.

Significance and Claims
The paper claims that while only a small fraction of light applied to the human body surface reaches the brain, the majority of energy loss occurs in the skin rather than the skull. This finding challenges the prior assumption that the skull is the primary barrier to transcranial therapy.

The authors conclude that:

1. Feasibility: Sufficient light energy can theoretically be delivered to the human brain in vivo if an appropriately designed delivery route and device are used.
2. Therapeutic Potential: By bypassing the highly attenuating skin layer (e.g., through surgical exposure of the skull), the light energy required for photo-oxygenation (based on a previously reported $ED_{50}$ of 6.05 J/cm²) could be achieved within approximately 2 hours of irradiation from the skull surface.
3. Methodological Contribution: The study provides a validated framework for estimating optical penetration depth across species by combining literature parameters with in vivo based modifications, offering a basis for defining device requirements and photocatalyst performance for AD therapy.

The authors maintain a modest tone regarding limitations, noting that probe invasiveness, the effects of anesthesia/euthanasia on blood flow, and the inability to separate scattering from absorption in transmittance-only measurements may lead to an underestimation of light delivery in the human simulation. However, they assert that these factors suggest the actual potential for light delivery may be even greater than simulated.