Theragen Looks to Data to Decode Spinal Fusion Recovery

Theragen presented new data from three studies at this month’s Society for Minimally Invasive Spine Surgery (SMISS) annual forum showing that recovery after fusion surgery with a bone growth stimulator can be predicted using patient demographics, health history, and surgical techniques.

The Jacksonville, Fla.-based company’s ActaStim S bone growth stimulator combines a therapeutic, non-invasive electrical signal with a modern, discreet design and data-rich digital platform. Its compact design is intended to support comfortable, discreet wear in everyday settings, potentially enabling longer daily use and supporting spinal fusion.

The digital platform’s charts combine automatically captured wear time and activity data with patient-reported pain scores, providing daily visibility into recovery and helping clinicians identify and address adherence issues remotely.

Study 1

The first poster examined predictors of pain after spinal fusion using SHAP analysis of longitudinal bone growth stimulator data.

Data showed that wear adherence was the strongest modifiable predictor of clinically meaningful pain improvement, with predicted gains increasing with daily wear time up to about 13 hours before plateauing. Baseline pain severity was the strongest predictor of achieving a minimal clinically important difference, while multilevel fusion and transforaminal lumbar interbody fusion (TLIF) were associated with a higher likelihood of predicted non-improvement, Theragen reported.

The study also determined that patients averaging under 13 hours of daily wear six weeks after surgery may benefit from additional adherence counseling or closer post-op monitoring.

Study 2

The second study assessed machine learning prediction of pain improvement after spinal fusion using longitudinal bone growth stimulator data.

Theragen’s XGBoost machine learning model outperformed standard logistic regression across all reported metrics, supporting nonlinear modeling for predicting clinically meaningful pain improvement, the company said. A standardized device-to-electronic medical record (EMR) pipeline achieved 99.2% completeness across approximately 195,000 patient-day observations, demonstrating the feasibility of integrating real-world BGS data at scale.

The model also achieved a 96.4% negative predictive value, possibly allowing earlier identification of patients unlikely to achieve meaningful pain improvement and informing post-op risk-stratification tools, pending external validation.

Study 3

The final study evaluated how a novel, granular longitudinal dataset can identify patient and surgical factors related to post-op spinal fusion pain.

Stratified clinical significance tiers (SCST) capture the magnitude and direction of longitudinal pain changes, offering a more nuanced alternative to a single pass/fail minimal clinically important difference (MCID) threshold, Theragen explained. Several patient and surgical factors were associated with post-op pain trajectories.

Degenerative disc disease, diabetes and renal disease were linked to improvement at six weeks, while anterior lumbar interbody fusion (ALIF) was associated with improvement at six months. Pseudoarthrosis, tobacco use, and spinal stenosis were associated with persistent pain as well.

The company concluded that despite statistical significance, effect sizes were generally modest, underscoring the multifactorial nature of post-op recovery.

Comments from Theragen and clinicians

Chris Mills, chief marketing officer: “ActaStim isn’t just collecting more data. It’s translating that data into guidance patients and providers can use. We’ve learned that no two spine surgery recoveries look the same. Many factors, including a patient’s age, health and surgical technique, can shape a patient’s path forward. Our Theragen care team can now help guide that path with real-world evidence.”

Ryan D. Snowden, MD, orthopedic surgeon specializing in spine surgery at the Tennessee Orthopaedic Alliance: “These findings further demonstrate the potential value of longitudinal, real-world data in helping clinicians better understand each patient’s recovery after spinal fusion surgery. Identifying meaningful patterns early may allow care teams to provide more individualized support and help patients remain engaged throughout the recovery process.”