Predicting the response to Lorazepam in catatonia: a network analysis approach.

BACKGROUND

Lorazepam, a widely used benzodiazepine, is the first-line treatment for catatonia. In case of resistance to Lorazepam, a trial-and-error approach is usually followed, increasing the risk of severe medical complications. To date, the mechanisms underlying non-response to Lorazepam remain unclear. The aims of this study were to determine whether network analysis can identify symptom-based predictors of response to Lorazepam and assess whether the structure of catatonic symptoms differs between responders and non-responders.

METHODS

A cohort of patients with catatonia (n = 136) was recruited at a specialized clinical center. Network analyses were conducted using the Bush-Francis Catatonia Rating Scale items, to describe the structure of catatonic symptoms in responders and non-responders to Lorazepam. Symptom clusters were identified using a spinglass algorithm and centrality measures were compared between groups. Additionally, network outcome analyses were performed to identify clinical predictors of Lorazepam response.

RESULTS

Three clusters of catatonic symptoms were identified, with similar structure between responders and non-responders. No significant differences were found in centrality measures. The most central symptoms were characteristics of hyperkinetic catatonia. Immobility / stupor emerged as the unique direct predictor of response to Lorazepam.

CONCLUSION

The overall structure of catatonic symptoms appears stable between responders and non-responders. Hypokinetic symptoms, particularly immobility and stupor, serve as strong predictors of Lorazepam response. Given the critical need for early treatment prediction to optimize therapeutic interventions and reduce catatonia-related mortality, further research is needed to refine clinical predictors of Lorazepam response.