Data-augmented machine learning refines the effective-concentration estimate for eculizumab in complement-mediated...

Area of Science:

• Pharmacology and Immunology
• Computational Biology
• Rare Disease Treatment

Background:

• Eculizumab effectively treats rare complement-mediated disorders but some patients show inadequate response.
• Current serum concentration targets for eculizumab are not based on robust pharmacokinetic-pharmacodynamic data.
• Uncertainty exists regarding the optimal eculizumab concentration for sustained complement C5 inhibition.

Purpose of the Study:

• To determine the serum eculizumab concentration required for complete C5 inhibition using machine learning.
• To evaluate the impact of synthetic data augmentation on the precision of pharmacokinetic-pharmacodynamic estimates.
• To reassess current eculizumab maintenance targets in rare diseases.

Main Methods:

• Digitized clinical data from four rare indications were used to train supervised machine-learning classifiers.
• Logistic regression on log-transformed eculizumab concentration was identified as the optimal model for predicting C5 inhibition.
• High-fidelity synthetic data were generated to augment real clinical data for improved model estimation.

Main Results:

• Machine learning models accurately predicted complete C5 inhibition based on serum eculizumab concentration.
• Synthetic data augmentation reduced the confidence interval for effective concentration by 92%.
• The estimated concentration for complete C5 inhibition was 310.8-335.2 µg/ml, potentially lower than current targets.

Conclusions:

• Current eculizumab maintenance targets for C5 suppression may be underestimated.
• Synthetic data augmentation can enhance the precision of pharmacological estimates in rare disease research.
• Prospective validation in independent cohorts is necessary before clinical application of these findings.