Treatment Selection and Optimization

1 Biomarker-Based Treatment Prediction

The integrated model (Chapter Integrated Multi-System Models) maps from biological parameters to treatment response. If a patient’s parameter values can be estimated from biomarker measurements, the model predicts which treatments will be most effective for that individual. The logic proceeds in three steps:

• Parameter estimation: Fit the model to the patient’s available data (metabolomics, cytokine panel, CPET, HRV, cortisol rhythm) to estimate the patient-specific parameter vector \(\hat{\mathbf{\theta}}_\text{patient}\)
• Intervention simulation: For each candidate treatment, modify the appropriate parameters (e.g., CoQ10 increases \([\text{UQ}]_\text{total}\), LDN modifies cytokine production rates) and simulate the model forward to predict symptom trajectories
• Treatment ranking: Rank treatments by predicted symptom improvement, accounting for parameter uncertainty through Bayesian credible intervals

This approach would enable rational treatment selection based on the patient’s dominant pathophysiological mechanism. A patient with primarily metabolic dysfunction (low \(\alpha_\text{CI}\), near-normal immune markers) would be directed toward mitochondrial support, while a patient with dominant immune activation (elevated cytokines, near-normal CPET) would be directed toward immunomodulation.

2 Dosing Optimization

For a given treatment, the model can predict the optimal dose by simulating dose–response relationships. The pharmacokinetic–pharmacodynamic (PK-PD) interface links drug plasma concentration \(C_\text{drug}(t)\) (from standard PK models) to parameter modifications in the pathophysiology model. For example, CoQ10 supplementation modifies \([\text{UQ}]_\text{total}\) as:

\[ [\text{UQ}]_\text{total}(t) = [\text{UQ}]_\text{baseline} + \Delta[\text{UQ}]_max \cdot \frac{C_\text{CoQ10}(t)}{K_\text{CoQ10} + C_\text{CoQ10}(t)} \tag{1}\]

where \(C_\text{CoQ10}(t)\) follows oral absorption and distribution kinetics. The model predicts the steady-state ATP improvement as a function of daily dose, identifying the dose above which additional supplementation provides negligible benefit (saturation) and below which the effect is indistinguishable from placebo.

3 Combination Therapy Design

The synergy analysis from Chapter Temporal Evolution and Disease Trajectories (Equation synergy) extends to systematic screening of treatment combinations. For \(n\) candidate treatments, the model can simulate all \(binom(n, 2)\) pairwise combinations and rank them by predicted synergy. Combinations targeting independent rate-limiting steps (e.g., ETC capacity + substrate supply + immune demand reduction) are predicted to show the greatest synergy, while combinations targeting the same bottleneck show diminishing returns.